Embed Size (px)
Citation preview
Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author.
- - ------------------ --
DEFOLIATION MANAGEMENT OF BIRDSFOOT TREFOIL (Lotus corniculatus L.)
A thesis presented in partial fulfilment of the requirements for the
degree of Doctor of Philosophy,
Institute of Natural Resources, Massey University, New Zealand
Walter A� 2001
This Thesis is dedicated to my wife Rossy, and our daughters Agustina and Bianca
for their love, help and support
ii
Abstract Hi
ABSTRACT
Ayalil, w. (2001). Defoliation management of bkd�root h'doil (Lolus comkul«I'U�
L.). rh.D Thesis, lnslltutc of �"ural R(,'ljo�tces: Mas-�ey niwr.sity, I'lIlml>rston
'(Irth. �w ZeaJand.
Birdsfoot trefoil ([.011$ corn �jjla1lls L.) i a lorage legume wide.\ cultivat,ed around lhe
\ orld, adapte.d to grow on infertile. drought-prone or acid soil, and wilJl a high feeding
value and bloat �fe forage. l'1owever. ja persistence is poor, limited by liIe
mallagemCJ1t 0 f d.cfolilllion and dist'�s:e inciden.cc. Adjustments i[l dcfo!iatioo strategies.
reproductive proct',s. I!. and JX>jJulation dynamic !,re seen as aU.em�liv, 10 il\crea<;e
production and persistence of birdsfoot trefoi l swarrl�. The objectives of this research
were to del ermine appropriate defoliation strategies for di ffere[lt bir-ds foot trefoi I
cultivars, in te,rms of tile frequency, inten ity and ti ming of defoli4ltion, and 10 quanti fy
morphological and physiological adaptaHOt\� and population dlanges in respanse lo
defolialiOll. A serie.::; or lhn::. lidd "!lcl one gl"!i.o;bQu,� e"p.rim�nL� "'.;re c(]nduc-ted in
M assey U oiversity. I"IIlmcrston onh, cw Zealand ([aliiudc 4{)"23 '8) and fNiA
Trelnta y Tres, Uruguay (tal.itll(Je 33�S4' S) from 1997 to 2QOO. lbe cultivars evaJuat.ed
were San Gabriel (Brazil), TNIA Dl1Ioo ( ruglla)'). Gra-5S1'lOds Goldie (New Zealand)
and t.eadfast (USA). ana-gemen( varied itl intensity of defoliation from 2, to IOcm
height, iA fre.quency tom 20 to 40 days, "nd ii'! timing Ule start of defoliatioA in dIe tirst
year from ve11,ctative to lau: mature stages. Also. combin lions of rest periods In
l'ltltl/mn, winter and slimmer were studied on pure and m.ixed bi.r<lsfoot trefoil swards.
A pte[irnimu)' snQn I.crm sl,udy with Gmsslands Goldic LA �w Z<:aland, showed that
ha,d defolia�ion (2 cm) iA spring reduced birdsfoot trefoil spring production {17%} alld
plant p pvlfltion (21%) compared wilh me average of la.:ter defohation (6 and 10 cm).
Root mass, crown ma, ,prilll\lry ,,[\(1 t\)l� I nU!1l her of hoou m2 find toot reserve�, ere
a 11 reduc�d under hard defoliation. Early autumn =t (last cut in April) impro\'cd phHlI
roO! reserve� and increased pring herbage· prOOllctiQ.Il (17%). TIle effects of intensity of
detolialion were �,)(mjirmed under controlled glassholl� conditions, wbere lax regime�
;.
16 """ 10 cm) iI .. , 'n', h<rtIog< pn><Io<t .................... dcf-"" (2,,) "he>!
okfoI� • � dIoy.....,,� I.";'" """"" •. � -.cl frequmI dorl'oliao_ r. cm.:!O
doys) "tu ,," pc' """ ..... pIaIII ...... , .... """ Iu _ .... fi",,, .. IIofolio<ion (Ill
....... doyo) ......... ;" Iw;,. er -. by ....... ,-. 'M .... L ....... ..... _
..... pIono ............. ill "' .... '« 1""'11> -. leaf ...... """me: leaf .... ..... • �.,
01''''',-.. ptO' piano ...... _......p '" ............. ,. "" 1h< ....... ,-. ..... fplant " ......
.......... .. ',.... dcl'ol ....... (2 ""'� no..fI«u of IktQI ... ion """""Y ;"m' ," ova ,imo,
r«h>:NIi """"" ''It. """ ""' ... toOl d� """ ..,.. f<'Wr>''' of toirdif_ ".fo;1
pl..,U.
Cul'i ...... <""I,..It<! ill • "'0 y..,.. .. ud� in UNguay difftml In piMI "",,i. (from .. nO·
�. "' ....., ')1>0'1� ";,,,., """"Y lIoCI' ....... ..."",.,. I)'peo) ond -tlool.Jsy,
Tb< ...... u ","',..-. .. cv. Slcadfasl \ou <>boom'.,.! ill I� or ........ . . �
..-. cu ...... s... Cp' Od """ 1l'1,\ 0.- .. .... 2.6 ..... l.J ,,_ -.: pc '"..;,.
_inIrodun ... '-NtwZn':'" _USA ...... r ... ono! .... ' '_ rcopocomy.
Tho: ..... of""�I ... ..-_"<d .......... t ... ofRnctqllllily.>Y,_._
S90-1lG .'la OM .... .,...itOli<y of "'PI;" ........ 2,·)9.'4 OM foul ........ l.IO-
_ A IIM b """,, _ ...... fiIn ..... lHI N DM "" "",4' "d ",. In
"""'-' "·,,h Ift"""" ..,..,..... habog< pc 1 _ ..... � ....... to. pi" .. <l<IOIUuoI ••
tm ........ ..... . Scmdurint: 1h< rlnl _ ..... tbmo ..... !'KI d,ff"" .... in .... , .• md
y .... COO"" ... ,,'n" 1If<';""" nr<Om<I1l> .,= ."ri_1< to . . . "ended <lc:foIi .....
interval (�O clay.) onc! ..... .,.,w. of 'J'II'If'»ima«ly 6 mO,,, .. In IM Y'" thal .11ow«!
pIonu 10 ....... 11d �i1> ....... ' .. f", ou« ... "" "'11"""",
"""'" _ .. in """"""tion ..... 111 .. bile d ..... ,.. _roo. \tOr,,;1 jk' t ... im " ....
.......... od by ... __ ODd Iu KJ1'<inI (10 ctn� """" -, ..... .- by
.......... .,.....(4 ctn)ond by.nl<p<S!boo.. d ... _I>< ... . a' 9 -' 11 .......
_.WI .J ...... _ .. Iom: _ ...... 6 ........ _. S' ==..,cll .. •• : ,od
...,.; � allNod,Mo Ir<'IOil odoimrc 11110 ........ ....wrr.' if ............
_ obo -. I�. _ire ......... 0« froroo "';1 oeod ___ ..... only bot,
S·I)� ....J« �rc ronoIiIiom '*'"'" .. ""'" ..,.; ...-. oInn&o..tirc oM, • ......!
...... ,."'� ... ...... ku: .. it>o:nas< r«:Nitonml of"... in:!i._lt.
Abstract
TIle results of these srudies were used to define practical managemeflt strategies to
oplimisc the production and p'-rsistence of birdsfO(}t trefoil swards, and plant
characlcri ·ties appropriate 10 Uruguayan nd New Ze.aland conditions.
Keywords: Lorlls cornf(�I/Ja1Us L.; birdsfoot trefoi I; CUlli vars; de foliation management;
forage production; nutritive value; persistence; plant morphology; caJ1bohydrate root
reserves; seed production.; soi I seed re er .... es; seedling cmergcn(X.
Acknowledgements
--.----------------------------
ACKNOWLEDGEMENTS
vi
There are many people that I would like to express my gratitude for their support and
contribution to the completion of this project.
I wish to thank sincerely to my chief supervisor, Professor John Hodgson (Institute of
Natural Resources, Massey University) for his continuous encouragement, guidance and
support over these years, providing a fertile environment of discussions and
development of research skills. His accurate supervision from the start to the end, made
possible to organize a split Ph.D program between New Zealand and Uruguay, and
contributed to extend more the research links between both countries.
Special thanks to my co-supervisor, Dr. Peter Kemp (Institute of Natural Resources,
Massey University) for provides always constructive comments, guidance and patience
throughout this study. His fine sense of humor made always the difference between
peaks of frustration and motivation.
My particular appreciation to my co-supervisors, Professor Ing. Agr. Milton Canimbula
(National Institute of Agricultural Research of Uruguay) and Ing. Agr. Diego Risso
(Head of Pastures Program of the National Institute of Agricultural Research of
Uruguay), for their generous help and suggestions particularly over the trials ran in
Uruguay, making this study productive.
I am extremely grateful to the National Institute of Agricultural Research of Uruguay
for the support, research faci lities and financial assistance, especially to the following
members: Mr. lP. Hounie and Ing. Agr. P. Bonino (past and present presidents of the
INIA Board and in their name to the other Board members), Dr. E. Indarte (National
Director), Dr. G. Cerizola (Human Resources) and Ing. Agr. H. Duran (Leader of
Animal Production group). To Dr. F. Montossi (Head of Sheep Program of the National
Institute of Agricultural Research of Uruguay) for his encouragement and help in many
stages of this project.
Acknowledgements vii
I would l ike to express my recognition to New Zealand, through the Ministry of Foreign
Affairs and Trade (MF A T) for provision of financial support by a NZODA scholarship.
My gratitude to Mr. Mac Herrera, consul of New Zealand in Uruguay, who provided
early arrangements for this scholarship. Also to the staff of the International Students
Office in Massey University for the assistance provided to me and my family.
My thanks to the staff of Natural Resources Institute, and particularly to Pastoral
Science Group at Massey University for their friendship and support over these years.
The assistance and invaluable criticism provided by Dr. C . Matthew, Mr P. Matthews
and Dr I. Valentine in different stages are well recognized. In particular, to Mr M.
Alexander, Mr R. Dissanayake, Ms K. Hamilton, Ms H. Kennedy, R. Levy, !. Manley,
Mr M. Osborne and others. Also to the staff of the Plant Growth Unit, Mr C. Johnstone,
L. Taylor, L. Sylva and H. Logan, and late Charlie Howell of Deer Research Unit of
Massey University. Very special thanks to Dr D. Woolley and Mr C. Rawlingson for
provide assistance to conduct the root carbohydrate analysis.
My recognition to those people that provided seeds of evaluated cultivars, to Dr. W.
Williams (Agresearch, Palmerston North, New Zealand), to Dr P. Beuselinck (USDA
ARS, Columbia, Missouri, USA) and lng. Agr. J. Bologna (Agronomy Faculty,
Uruguay).
My sincere thanks to lng. Agr L. Helguera and lng. Agr. G. Zorril la, past and present
Directors of INIA Treinta y Tres, to provide the best technical and human conditions to
execute this project. To the staff of the "old" Estaci6n Experimental del Este (IN lA
Treinta y Tres Research Station, Uruguay) for their assistance, warm support and
encouragement to achieve this goal .
I would like to acknowledge the generous field and laboratory assistance given by Mr
C. Carmona, Mr G. Ferreira, Mr J. Jackson and Mr N. Serron of Pastures Group of
INIA Treinta y Tres, during the trials conducted in Uruguay. Their attention and
accuracy during many hours that we shared running these experiments will be never
forgotten. Many thanks goes to my colleague of Pastures group lng. Agr. R. Bermudez,
Acknowledgements viii
that gave me unconditional assistance. Thanks to the students Mr. C. Machado and Mr.
I. Nuiiez from the Agronomy Faculty of Uruguay, who co-laboured with me collecting
and sharing information.
The friendship and encouragement of my fel low graduate students at Massey University
provided an excellent environment, especially to Dr. S. Assuero, Dr. S. Bluett, Mr. W.
Beskow, Mrs. R. Briant, Mr D. Cabrera, Ms. D. Carvalho, Dr. N. Devkota, Mrs M.
Ercolin, Dr. A. Guevara, Mr. C. Hepp, Mr. F. Hughes, Mr. and Mrs. Laborde, Dr. G. Li,
Mrs. H. Logan, Dr. I . Lopez, Dr. Z. Nie, Dr. S . Oppong, Dr. C. Poli, Dr. D. Real, Mrs
C . Realini, Mr. A. Romera, Mr. 1. Rossi, Mrs P. SalIes, Mr. T. Pande, Mr. A. Wall and
others.
My sincere thanks to my parents and sister for their education, support and continuous
encouragement from early steps at the primary school (Escuela Rural No. 84 de Molles
de Gutierrez, Lavalleja, Uruguay) to nowadays. The support of our friends, in special to
my relations in Uruguay, is greatly appreciated.
Finally I want to express my gratitude to my wife Rossy and daughters Agustina and
Bianca. Their technical help, tolerance, sacrifice and motivation throughout these "long
and hard" years, accepting the challenge to complete this project, contributed to make
this study successful.
WaIter Ayala
April 5, 200 1
Table of Contents ix
TABLE OF CONTENTS
IlIJ�TFlt\CT........................................................................... iii
IlCKNOWLEDGEMENT�.... .............. ............. ...................... ... vi
T IlIJLE OF CONTENT�... .. ... ..................... ................ .............. ix
LI�T OF T IlIJLE�................................................................... xiv
LI�T OF FIGURE�...... ....... .............. .................................. ..... xix
LIST OF PLAT,ES . ..... . " ...... """ .. ,, ....... ,."" .... """ .. """" .... " .. """ .. "" .. """" .. ,, .... ... xxv
1. GENERAL INTRODUCT.ION .. " ... """" .. ,, ............ " .. " ........ " .. ".... ...... 1
1.1 INTRODUCTION............................................................ 2
1.2 OBJECTIVES ................................................................ 4
1.3 THESIS OUnINE...................................................... ..... 5
1.4 REFERENCES................................................................ 6
2. REVIEW OF THE LITERATURE......... .................. •... ..... ...... 9
2.1 INTRODUCTION ............................. , ...... '" .... ,. ... ...... ...... 10
2.2 AGRONOMIC CHARACTERISTICS.................................... 11
2.2. I Plant morphology. . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . ... .. 11
2.2.2 Birdsfoot trefoil habitat............................................. 1 2
2.2.3 Herbage production............................................. ..... 1 3
2.3 NUTRITIVE VALUE................................................. ....... 1 4
2.3 .1 Digestibility .............................. " . .. . . . . . . . . . .. . . . .... .. . .. 15
2.3.2 Crude protein..................... .... .................. ... ........... 1 5
2.2.3 Fibre content.................................................. ... . .... 16
2.2.4 Mineral content................................. ............. ......... 16
2.2.5 Condensed tannins content.......................................... 17
2.4 UTILIZATION IN FARMING SySTEMS............................... 18
2.4.1 New Zealand........................... .............................. 18
Table of Contents x
2.4.2 Uruguay ....................... , .......... .... ........... ..... ..... .... 19
2.5 DEFOLIA nON MANAGEMENT................................. ....... 20
2.5.1 Frequency and intensity of defoliation and growth habit...... 2 1
2.5.2 Leaf area index and photosynthetic efficiency.... . . . . . . . . .. ... 22
2.5.3 Carbohydrate reserves......... .... ... ....................... ...... 2 3
2.6 DISEASE INCIDENCE.................................................. 2 4
2.7 PERSISTENCE REVISITED............................................. 25
2.7.1 Plant strategy........................................................ 25
2.7.2 Improved persistence approach.... ....... ..... .................. 27
2.8 CONCLUSIONS............................................................ 30
2.9 REFERENCES.............................................................. 32
3. CHANGES IN THE MORPHOLOGY, PRODUCTION AND
POPULATION OF Lotus corniculatus L. cv. GRASSLANDS
GOLDIE IN RESPONSE TO SEASONAL DEFOLIATION
REGIMES ................................ ., ................................. ., . .,... 45
3.1 ABSTRACT................................. .............. . .................. 46
3.2 INTRODUCTION.......................................................... 47
3.3 MATERIALS AND METHODS.......................................... 48
3.3.1 Measurements....................................................... 49
3.3.1.1 Herbage production....................................... 49
3.3.1.2 Plant density, size and morphology..................... 49
3.3 .1.3 Root carbohydrate analysis ....... ......... . . ........... , 50
3.3.2 Statistical analysis...... ........ . ... ............... ........ ......... 5 1
3.4 RESULTS.................................................................... 5 1
3.4.1 Climate conditions during experimental period................ 5 1
3.4.2 Herbage production................................................ 52
3.4.3 Plant density....... ..................... . ............. ...... .... ..... 54
3.4.4 Plant morphology ................ , ................... '" ... ... ...... 5 5
3.4.5 Carbohydrate root reserves........................................ 57
3.5 DISCUSSION......... ... ... ... ...... ... ...... ... ....... ....... .......... ... 59
3.6 CONCLUSIONS............................................................. 62
3.7 REFERENCES............... ... ..................... ................... ..... 63
Table of Contents
4. EFFECTS OF DEFOLIATION INTENSITY ON GROWTH,
BIOMASS DISTRIBUTION, AND MORPHOLOGICAL AND
PHYSIOLOGICAL CHANGES OF BIRDSFOOT TREFOIL IN
xi
GLASSHOUSE CONDITIONS.............................................. 67
4.1 ABSTRACT............................................................... ... 68
4.2 INTRODUCTION........................................................... 69
4.3 MATERIALS AND METHODS.......................................... 70
4.3.1 Measurements........................................ ........... ..... 72
4.3.2 Statistical analysis.................................................. 73
4.4 RESULTS........................... ...................................... .... 75
4.4.1 Growth analysis..................................................... 75
4.4.2 Biomass production................................................ 76
4.4.3 Carbohydrate root reserves............................. ........... 79
4.4.4 Dynamics of plant components................................ ... 8 0
4.4.4.1 Below-ground components............................... 80
4.4.4.2 Above-ground components............................... 86
4.4.5 Relationships between herbage harvested and plant
components.............................. ............................ 90
4.4.5.1 The influence of below-ground plant components.... 90
4.4.5.2 Residual leaf area for regrowth........................... 91
4.4.5.3 Carbohydrate root reserves and regrowth........... .... 92
4.4.5.4 Effect of defoliation intensity on plant components.. 93
4.4.6 Plant survivaL......................................... ... ... ....... 94
4.5 DISCUSSION............................................................... 94
4.6 CONCLUSIONS............................................................ 98
4.7 REFERENCES.............................................................. 99
S. PERFORMANCE OF FOUR Lotus corniculatus L. CULTIV ARS
IN RESPONSE TO INTENSITY AND TIMING OF
DEFOLIATION......... ................. ................. ..... ...... ....... ..... 102
5.1 ABSTRACT................................................................. 103
5.2 INTRODUCTION.......................................................... 104
Table of Contents xii
5.3 MATERIALS AND METHODS........ ....... ........ .......... ... ...... 106
5.3.1 Measurements....................................................... 108
5.3 .1.1 Forage production. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... 108
5.3 .1.2 Forage quality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
5.3.1.3 Condensed tannins......................................... 109
5.3.1.4 Plant density and morphology............................ 109
5.3.1.5 Plant architecture.......................................... 110
5.3.2 Statistical analysis.............................................. .... 110
5.4 RESULTS........................... ............ ........................ ...... 112
5.4.1 Climate conditions during experimental period................ 112
5.4.2 Herbage production to first harvest.................. ............ 113
5.4.3 Annual herbage production ................. . ......... '" ...... '" 1 1 6
5.4.4 Growth rates........ ..... ..... ... ...... ........ ... ....... ... ..... .... 118
5.4.5 Sward structure characteristics............................... ..... 124
5.4.5.1 Residual forage................................. ............ 124
5.4.5.2 Plant architecture.......................................... 126
5.4.6 Nutritive value.................................................. ..... 128
5.4.6.1 Nutritive value of accumulated BFT forage............ 129
5.4.6.2 Nutritive value parameters under regular defoliation. 132
5.4.6.3 Condensed tannin content................................. 135
5.4.7 Plant density.............. .... ............... ......... ...... ......... 135
5.4.8 Plant morphology... ................................. ... ....... ..... 136
5.4.8.1 Integrated morphology analysis............... .......... 138
5.5 DISCUSSION............................................................... 144
5.5.1 Productivity and adaptation ofBFT to the eastern region of
Uruguay ......... " ....................... , ........... " . . . . . . . . . . . ... 144
5.5.2 Growth, plant type and defoliation management........ ....... 1 4 6
5.5.3 Forage quality.. . . .. .. .. . . .. . .. .. . . .. .. .. .. . .. . . .. . .. . .. . .. .. .. . . . . . 1 48
5.5.4 Forage accumulation ........... . ........................... .. ,..... 149
5.6 CONCLUSIONS ............... '" ..... , ..... ....... ...... ... ....... ........ 1 52
5.7 REFERENCES. ........ ............ ......... ... .... ........ .... ..... ... ..... 1 53
Table of Contents
6. FORAGE PRODUCTION AND PERSISTENCE OF
BIRDSFOOT TREFOIL (Lotus corniculatus L.) IN MIXTURE
WITH WHITE CLOVER IN RESPONSE TO DIFFERENT
xiii
STRATEGIES AND INTENSITIES OF DEFOLIATION............ 157
6.1 ABSTRACT..................... ..... . ....................................... 1 58
6.2 INTRODUCTION.......................................................... 160
6.3 MATERIALS AND METHODS ..... , ... ... .......... ... ... ..... .... ..... 162
6.3.1 Measurements........................... .................. .......... 164
6.3.2 Statistical analysis...... . ......... ..... ...... .......... ............. 165
6.4 RESULTS................................................ ........... . ........ 167
6.4.1 Climate conditions during experiment..................... ...... 167
6.4.2 Herbage production and quality......... ...... .................. 167
6.4.2.1 Pre-grazing and post-grazing sward height............ 167
6.4.2.2 Annual and seasonal herbage accumulation ........... 168
6.4.2.3 Species contribution... .................................... 170
6.4.2.4 Herbage quality............................................ 174
6.4.3 Plant density......................................................... 177
6.4.4 Plant morphology................................................... 179
6.4.4.1 Primary shoots......................................... ..... 179
6.4.4.2 Secondary shoots.. ... . . ........... .......... .......... .... 180
6.4.4.3 Crown mass...... .... ...... .... .... .............. .......... 181
6.4.4.4 Root mass................................................... 182
6.4.4.5 Root diameter ............................................. 183
6.4.5 Seed production..................................................... 184
6.4.5.1 Seed yield components.. .... .......... ..... ......... ...... 185
6.4.5.2 Patterns of seed production ..... . ...... . . ................ 187
6.4.6 Seed soil reserves..................... .................... .......... 189
6.4.7 Seedling emergence... . . ................ . ... ................. ....... 191
6.5 DISCUSSION.................. ............. . ............................... 195
6.6 CONCLUSIONS............................................................. 199
6.7 REFERENCES .............................................................. 200
------------------
Table of Contents xiv
7. INTEGRATING DISCUSSION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
7. 1 INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
7.2 BIRDSFOOT TREFOIL GENOTyPES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
7.3 DEFOLIATION MANAGEMENT, PRODUCTION AND PLANT
SURVIVAL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
7.4 THE ROLE OF THE SOIL SEED BANK ON POPULATION
DYNAMICS AND PERSISTENCE OF BIRDSFOOT TREFOIL. . . 2 1 3
7 .5 PRACTICAL MANAGEMENT RECCOMENDA TIONS . . . . . . . . . . . 2 1 6
7 .5 . 1 Seasonal management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 7
7 .5 .2 General management. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 9
7.6 CONCLUSIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
7.7 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
APPENDIX I . METHOD TO MEASURE TOTAL A V AILABLE
CARBOHyDRATES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
APPENDIX n. METHOD TO EV ALUATE SOIL SEED
RESERVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
List of Tables xv
LIST OF TABLES
Table 3- 1 . Monthly rainfall and soil temperature at 1 00 mm during the
evaluation period and the 60-year average (Source:
Table 3 -2.
AgResearch, Palmerston North) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1
The effect of defoliation management on BFT accumulation
(kg DM/ha), growth rate (kg DM/ha/day) and contribution to
total production (%) during spring. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Table 3-3 . Leaf/(Leaf+Stem) ratio (dry weight) in BFT under different
defoliation frequencies and intensities in spring. . . . . . . . . . . . . . . . . 53
Table 3-4. Effects of defoliation treatments on plant morphology
parameters/m2 and plant density in December 1 997 for
Table 4- 1 .
Table 4-2.
Table 4-3 .
Table 4-4.
Table 4-5.
birdsfoot trefoil cv. Grasslands Goldie . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Cutting and destructive harvests schedule for BFT pots from
September to December 1 997 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1
Cumulative above-ground growth, above-ground biomass at
the time of destructive harvest and below-ground mass of
BFT cv. Grasslands Goldie under different intensities of
defoliation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Content of starch and free sugars in roots of BFT cv.
Grasslands Goldie under different intensities of defoliation
over 1 20 days (actual values expressed in mg/plant and log
values used for statistical analysis) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Crown mass, primary and secondary roots mass and root
diameter (expressed in natural log (x+ l ) values) of BFT cv.
Grasslands Goldie managed at cutting heights of 2, 6 and 1 0
cm and one undefoliated control treatment, over 1 20
days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
Primary and secondary shoots per pot (actual and log values)
of BFT cv. Grasslands Goldie managed at cutting heights of
List of Tables
Table 4-6.
Table 4-7.
Table 4-8.
Table 5- 1 .
Table 5-2.
Table 5-3 .
Table 5 -4.
Table 5-5.
Table 5-6.
Table 5-7.
-------------- - -- -
2, 6 and 1 0 cm and one undefo liated control treatment, over
xvi
1 20 days. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
Leaf area ratio (LAR), specific leaf area (SLA), weight per
leaf (wt. !leat) and leaves/gram of birdsfoot trefoil under
different defoliation intensities over time . . . . . . . . . . . . . . . . . . . . . . . . 89
Regressions between BFT herbage harvested (Y, g DM/pot)
and below-ground parameters during two periods of growth
after defoliation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1
Multivariate analysis of variance (Manova test) performed
on morphological components affected by defoliation height
at three successive destructive harvests (40, 80 and 1 20
days). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Soil characteristics at experimental site in Palo a Pique
Research Unit. (Source: Laboratory of Soils, INIA La
Estanzuela, Uruguay) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cutting schedule of BFT plots sown on May 1 998 . . . . . . . . . . . . .
Monthly rainfall and soil temperature at 50 mm depth during
the evaluation period and the 8-year average (Source: INIA,
1 06
1 07
Treinta y Tres, Uruguay) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 2
Herbage harvested (kg DMlha) and plant height at harvest
(cm) of local BFT cultivars from sowing (8/5/98) to three
different stages of development in the establishment year. . . . .
Herbage harvested (kg DM/ha) and plant height at harvest
(cm) of introduced BFT cuItivars from sowing (8/5/98) to
three different stages of development in the establ ishment
year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . .
The effect of defoliation intensity applied in Year 1 and 2 on
annual herbage production (t DMlhalyear) in Year 2 of two
1 14
1 1 5
groups of BFT cultivars (local and introduced). . . . . . . . . . . . . . . . . . 1 1 8
Effects of Year 1 and Year 2 treatments on overall growth
rates (kg DMlha/day) during spring-summer of Year 2 in
two pairs ofBFT cultivars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 24
�-------------- - -� -
List of Tables xvii
Table 5-8. Residual herbage mass, residual leaf area index (LAI) and
leaf/(leaf+stem) ratio after cutting of local cultivars managed
at two defoliation heights and three times of initial
Table 5-9.
defoliation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 25
Residual herbage mass, residual leaf area index (LA I) and
leaf/(leaf+stem) ratio of introduced cultivars managed at two
defoliation heights and three times of initial defoliation . . . . . . . 1 26
Table 5- 1 0. Nutritive value of local BFT cultivars for the first harvest
from sowmg to three different periods of dry matter
accumulation for plots defoliated at two different heights . . . . . 1 30
Table 5 - 1 1 . Nutritive value of introduced BFT cultivars for the first
harvest from sowing to three different periods of dry matter
accumulation for plots defoliated at two different heights . . . . . 1 3 1
Table 5 - 1 2. Nutritive value of local BFT cultivars managed at two
defoliation heights over 40 day intervals from November
1 998 to April 1 999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 33
Table 5- 1 3 . Nutritive value of introduced BFT cultivars managed at two
defoliation heights and 40 day intervals from December
1 998 to April 1 999. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 34
Table 5 - 1 4. Condensed tannin (CT, g/kg DM) contents of four BFT
cultivars at vegetative stage in early spring. . . . . . . . . . . . . . . . . . . . . . 1 35
Table 5 - 1 5 . Morphological parameters of four BFT cultivars 1 1 0 days
after a sowing made on May 8, 1 998. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 37
Table 5 - 1 6. Individual plant morphology parameters in BFT cultivars
affected by the intensity and timing of defoliation at the end
of the first year (March 1 999) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 40
Table 5- 1 7 . Individual plant morphology parameters in BFT cultivars
Table 6- 1 .
affected by the intensity and timing of defoliation at the end
of the second year (March 2000). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 42
Soil nutrients levels at experimental site in Palo a Pique
Research Unit. (Source: Soils Laboratory, INIA La
Estanzuela, Uruguay) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 62
List of Tables xviii
Table 6-2. Seasonal average heights (cm) of pre and post-grazing
herbage (showed as pre and post) and standard deviation
Table 6-3 .
Table 6-4.
Table 6-5.
Table 6-6.
Table 6-7.
Table 6-8.
Table 6-9.
values in brackets for all treatment combinations. . . . . . . . . . . . . . . 1 68
Annual and seasonal herbage accumulation (kg DM/ha) of
an oversown birdsfoot trefoil/white clover mixture managed
under different strategies and intensities of grazing during
the third and fourth year after establishment. . . . . . . . . . . . . . . . . . . . . 1 69
Annual and seasonal herbage accumulation of birdsfoot
trefoil (kg DM/ha) in a birdsfoot trefoil/white clover mixture
managed under different strategies and intensities of grazing
during the third and fourth year after establishment. . . . . . . . . . . . . 1 7 1
Annual and seasonal herbage accumulation o f white c lover
(kg DM/ha) in a birdsfoot trefoil/white clover mixture
managed under different strategies and intensities of
grazing during the third and fourth year after establ ishment. . 1 72
Annual and seasonal herbage accumulation of grasses (kg
DM/ha) in a birdsfoot trefoil/white clover mixture managed
under different strategies and intensities of grazing during
the third and fourth year after establishment. . . . . . . . . . . . . . . . . . . . 1 74
Seasonal averages of in vitro organic matter digestibility
(g/kg DM) of a birdsfoot trefoil/white clover oversown
mixture, during two years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Seasonal averages of nitrogen content (glkg DM) of a
birdsfoot trefoil/white clover oversown mixture, during two
years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Seasonal averages of acid detergent fibre (g/kg DM) of a
birdsfoot trefoil/white clover oversown mixture, during two
1 75
1 76
years . . . . . . . . . . . . '" . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 77
Table 6- 1 0. Number of primary shoots/plant of BFT under different
strategies and intensities of defoliation from April 1 998 to
December 1 999. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 80
---- --- ------------------ - - -
List of Tables
Table 6- 1 1 . Evolution of secondary shoots (no.lplant) of BFT plants
under a combination of four defoliation strategies and two
xix
defoliation intensities from April 1 998 to December 1 999 . . . 1 8 1
Table 6- 12 . Evolution of crown mass (g/plant) of BFT plants under four
defoliation strategies and two defoliation intensities from
April 1 998 to December 1 999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 82
Table 6- 1 3 . Evolution of root mass (g/plant) of BFT plants under four
defoliation strategies and two defoliation intensities from
April 1 998 to December 1 999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 83
Table 6- 1 4. Evolution of root diameter (mm) of BFT plants under four
defoliation strategies and two defoliation intensities from
April 1 998 to December 1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 84
Table 6- 1 5 . Annual seed production (g/m2) of birdsfoot trefoil (BFT)
and white clover (WC) in mixture under different strategies
and intensities of defoliation during two years . . . . . . . . . . . . . . . . . . 1 85
Table 6- 1 6. Inflorescences/m2 (I), viable seeds/m2 (S) and 1 000 seed
weight (W) (g) of BFT/WC mixture under different
strategies and intensities of defoliation, evaluated during
two years . . . . . . . . . . . . " . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 86
Table 6- 1 7. Soil seed reserves and 1000 seed weight parameters in
mixed birdsfoot trefoil (BFT) and white clover (WC)
swards under different defoliation strategies and intensities,
during two years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 9 1
Table 6- 1 8. Seedling emergence (no./m2) and percentage of emergence
from soil seed reserves of birdsfoot trefoil (BFT) and white
clover (WC) under different strategies and intensities of
defoliation from March - August 1 999 . . . . . . . . . . . . . . . . . . . . . . . . . . 1 94
Table 7- 1 .
Table 7-2.
Description of experiments conducted in this project. . . . . . . . . . .
Summary of published research with emphasis in defoliation
intensity on birdsfoot trefoil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
205
209
List of Figures
LIST OF FIGURES
Figure 2- 1 . Theoretical model for population dynamics of BFT (partially
adapted from Emery et al., 1 999). Developmental stages are
seed (nl ), seedling (n2), mature vegetative plant (n3) and
mature reproductive plant (n4). The references al to 4 represent
xx
transitional stages from stages n l to n4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Figure 2-2. Research priorities in birdsfoot trefoil . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1
Figure 3 - 1 . Treatment effects on seasonal changes in birdsfoot trefoil
plant density. Vertical bar represents SEM, (n= 1 6) . . . . . . . . . . . . . 54
Figure 3-2. Effects of defoliation management on secondary shoots per
plant of BFT in December, under two intervals and three
intensities of defoliation. Vertical bar represents SEM,
(n==8) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Figure 3-3 . Carbohydrate reserves in BFT roots in early spring of plants
receiving two contrasting autumn managements (early rest in
April or late rest in June). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 8
Figure 3 -4. Carbohydrate reserves in BFT roots at the end of spring
(December 1 997) of plants managed under two intervals and
three intensities of defoliation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 8
Figure 3-5 . Influence of defoliation height on plant popUlation, shoots
density and herbage contribution to total pasture production
in BFT cv. Grasslands Goldie in spring. . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1
Figure 4- 1 . Accumulation rate of herbage harvested (DM g/pot/day) of
BFT cv. Grasslands Goldie defoliated at three intensities in
controlled conditions. Vertical bars represent SEM, (n=
number of observations for each treatment mean, were 1 5,
1 0, 1 0, 5 , 5 and 5 for 0-20, 20-40, 40-60, 60-80, 80- 100 and
1 00- 1 20 day intervals respectively) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
---------------------- � --
Figure 4-2. Relative growth rate (DM gig/day) of BFT cv. Grasslands
Goldie defoliated at three intensities in controlled conditions.
Vertical bars represent SEM, (n= number of observations for
each treatment mean, were 1 5, 1 0, 1 0, 5 , 5 and 5 for 0-20,
20-40, 40-60, 60-80, 80- 1 00 and 1 00- l 20 day intervals
respectively) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Figure 4-3. Crown mass (g DM/pot) of BFT cv. Grasslands Goldie at
cutting heights of 2, 6 and 10 cm and for an undefoliated
control over 1 20 days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Figure 4-4. (a) Primary roots mass and (b) secondary roots mass (DM
g/pot) of BFT cv. Grasslands Goldie at cutting heights of 2,
6 and 1 0 cm and for an undefoliated control over 1 20
days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1
Figure 4-5 . Root diameter (mm) of BFT cv. Grasslands Goldie at cutting
heights of 2, 6 and 1 0 cm and for an undefoliated control
over 1 20 days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Figure 4-6. Regression between residual leaf area and herbage growth of
BFT during three periods (day 0 to 20, 40 to 60 and 80 to
1 00) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Figure 4-7. Regression between total carbohydrate root reserves and
BFT herbage harvested of plants defoliated at 20 day
intervals (data comprised average/treatment between day 40
to 60 and 80 to 1 00) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Figure 5- 1 . Effective rainfall, evapotranspiration (EVP) and soil water
balance for the period April 1 998-February 2000 in Palo a
Pique, Research Unit (Raid Bermudez and Jose Terra,
personal communication). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 3
Figure 5-2. Annual herbage production (t DM/ha/year) of BFT cultivars
affected by timing of initial defoliation during the Year 1 .
Vertical bars represent SEM for each group of cultivars,
(n=8). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 6
List of Figures
Figure 5-3 .
Figure 5-4.
Figure 5-5 .
Seasonal variation in growth rates (kg DM/halday) of BFT
cultivars (a) INIA Draco, (b) San Gabriel, (c) Steadfast and
(d) Grasslands Goldie under two defoliation intensities.
Vertical bars represent SEM; *, (P<0.05); **, (P<O.O I ) and
NS, (differences not significant) for corresponding growth
periods. The number of observations (n) for each treatment
mean was 4 (4 Nov- 1 5 Dec), 8 ( 1 5 Dec-25 Jan), 1 2 (25 Jan-8
March) and 1 2 (8 March- I S Apr) for INIA Draco and San
Gabriel. For Steadfast and Grasslands Goldie (n) was 4 (4
Dec- 1 3 Jan), 8 ( 1 3 Jan-22 Feb) and 1 2 (22 Feb-5 Apr) . . . . . . . .
Effect of height of defoliation on growth rates of local BFT
cultivars from November 1 998 to April 1 999. (**, P<O.O I ;
NS, not significant; SEM, standard error of the mean, n=8 (4
Nov- 1 5 Dec), n=1 6 ( 1 5 Dec-25 Jan), n=24 (25 Jan-8 March
and 8 March- I S Apr) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Growth rate of introduced BFT cultivars defoliated at two
different heights from December 1 998 to April 1 999. (* * ,
P<O.O I ; NS, not significant; SEM, standard error of the
mean), n=4 (4 Dec- 1 3 Jan), n=8 ( 1 3 Jan-22 Feb), n=1 2 (22
xxii
1 20
1 2 1
Feb-5 Apr» . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 1
Figure 5-6. Growth rates of (a) local cultivars and (b) introduced
cultivars of BFT, influenced by the number of previous
defoliation during spring and summer. (**, P<O.O I ; NS,
differences not significant for corresponding growth periods;
SEM, standard error of the mean, n= 1 6). Bars with same
colour represent the same treatment across periods . . . . . . . . . . . . . 1 22
Figure 5-7. Vertical distribution of tissues in BFT swards strata for local
cultivars under two defoliation heights in December 1 999,
determined from inclined point quadrat contacts . . . . . . . . . . . . . . . . 1 27
Figure 5-8. Vertical distribution of tissues in BFT swards strata for
introduced cultivars under two defoliation heights in
December 1 999, determined from inclined point quadrat
contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
-------------��-- -
Figure 5-9. Changes in plant density of four BFT cultivars over the
period July 1 998 - March 2000. Numbers in brackets
indicate the SEM between treatments at corresponding dates,
(n=24) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 36
Figure 5 - 10. Canonical analysis for morphology parameters in the Year 1
(a) and Year 2 (b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 43
Figure 5 - 1 1 . Changes in dry matter harvested (DM), digestible organic
matter harvested (DOMA) and organic matter digestibility of
two BFT cultivars (a) San Gabriel and (b) INIA Draco from
sowing to three physiological stages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 47
Figure 6- 1 . Description o f grazing strategies applied on the birdsfoot
trefoil-white clover mixture from April 1 998 to March 2000.
Each strategy was defoliated at 4 and 10 cm height. . . . . . . . . . . . . 1 63
Figure 6-2. Changes in BFT density (adults plants) under (a) four
defoliation strategies and (b) two defoliation intensities from
April 1 998-March 2000. Vertical bars indicate SEM
(nstrategies= 1 6, nintensities=32), and NS not significant
differences at corresponding sampling dates. . . . . . . . . . . . . . . . . . . . . 1 78
Figure 6-3. Patterns of seed production in BFT (g/m2) over two summer
seasons affected by defoliation strategies (a) and by
defoliation intensities (b). **, P<O.O I ; *, P<0.05; NS, not
significant; numbers in brackets, SEM (na=1 6, nb=32) . . . . . . . . . 1 88
Figure 6-4. Patterns of seed production in WC (g/m2) over two summer
seasons affected by defoliation strategies (a) and by
defoliation intensities (b). **, P<O.O l ; *, P<0.05; NS, not
significant; numbers in brackets, SEM (na== 1 6, nb=32) . . . . . . . . . 1 89
Figure 6-5. Seedling emergence patterns of (a) birdsfoot trefoil (BFT)
and (b) white clover (WC) and (c) climatic parameters,
evaluated on field from June to December 1 998 . . . . . . . . . . . . . . . . 1 93
List of Figures
Figure 7- 1 . Results related to the effect defoliation intensity in BFT
growth from information presented in (a) Chapters 3 and 4,
xxiv
(b) Chapter 5 and (c) Chapter 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 2 1 0
List of Plates
Plate 3- 1 .
Plate 3-2.
Plate 4- 1 .
Plate 4-2.
Plate 5- 1 .
Plate 5-2.
Plate 6- 1 .
Plate 6-2.
----------------- -
xxv
LIST OF PLATES
Birdsfoot trefoil plant showing the morphological
parameters evaluated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Number of plants of birdsfoot trefoil in December 1 997
under three intensities of defoliation (2, 6 and 10 cm).
Samples represent 250x250 mm quadrat. . . . . . . . . . . . . . . . . . . . . . . . . . 55
General view of BFT pots at the time of start (a) and during
the trial (b) an disease symptoms on some BFT plants
(c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ' " 74
Above and below-ground biomass of BFT plants defoliated
at different intensities and undefoliated control at 40 days
destructive harvest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
General view of BFT cultivars under different intensities and
timing of defoliation in the year of establ ishment (Year
1 ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 1
BFT plants (6 months old) of San Gabriel and Grassland
Goldie (a), and one year old plant of Steadfast showing the
development of rhizomes (b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 8
(a) Sheep grazing birdsfoot trefoil/white clover oversown
mixture in the experimental site at Palo a Pique Research
Station, Treinta y Tres, Uruguay. (b) Postgrazing sward
height were recorded to maintain contrasting defoliation
intensities of 4 and 1 0 cm. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 66
Seedling emergence was checked regularly from cores
placed in an adjacent area to experimental site and
maintained free of ground cover. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 92
Chapter One
1 . GENERAL INTRODUCTION
1 . 1 INTRODUCTION
1 .2 OBJECTIVES
1 .3 THESIS OUTLINE
1 .4 REFERENCES
. . ..
1
Chapter One
1 . 1 INTRODUCTION
------------------ - - - -- -
2
Lotus corniculatus L. (birdsfoot trefoil, BFT) is one of the most common and widely
distributed forage legumes around the world. It can be found in areas across Europe,
Asia, Africa, South America, North America and Australasia. BFT shows a high degree
of genetic variability that confers the capabilities of a wide range of adaptation and
distribution (Kirkbride, 1 999; Steiner, 1 999).
Approximately 4.5 million hectares of BFT are reported around the world, in mixed
pastures or pure stands for grazing or forage conservation, the main areas being
concentrated in Uruguay, USA, Argentina, Austria and Italy (Blumenthal and McGraw,
1 999). BFT agronomy has been extensively reviewed by MacDonald ( 1946), Seaney
and Henson ( 1970), Turkington and Franko ( 1 980), Grant and Marten ( 1 985), Jones and
Turkington ( 1 986), Blumenthal et al. ( 1 994), and Frame et al. ( 1 998) among others. It i s
particularly adapted to grow on low fertility, droughty or acid soils (Seaney and
Henson, 1 970) as a pioneer or alternative species, gaining the description of "lucerne of
poorest soils". Its feeding value is comparable with lucerne (Marten and Jordan, 1 979),
and it is a bloat safe forage (Blumenthal and McGraw, 1 999).
Over decades, the main problem to solve has been the persistence of BFT swards. Major
factors that determine BFT stand persistence are grazing management and diseases
incidence. BFT is sensitive to intensive and frequent defoliation, and when growing in
warm and humid environments it can only maintain dense and productive stands for 2-3
years. The development of new cultivars with improved disease resistance could solve
some of the reported problems, but this will not constitute by itself a genuine solution.
In practice, there is general acceptance to consider BFT as a short-lived perennial
legume under grazing (Pierre and Jackobs, 1 953) . A strategy to improve sward
persistence can be through natural reseeding of existing soil seed banks to replace plant
losses (Blumenthal and McGraw, 1 999), which is effective in areas where seed
production is not l imited (Beuselinck and McGraw, 1 989).
Chapter One 3
BFT is used in pastoral systems of Uruguay for grazing in pure or mixed swards with
grasses like annual ryegrass (Lolium multiflorum), tal l fescue (Festuca arundinacea) or
cocksfoot (Dactylis glomerata), and legumes l ike white clover (Trifolium repens).
Reduced stand persistence is reported, with declining herbage production and
population after the second year (Formoso, 1 993; Altier, 1 997). Recent attempts to
overcome these limitations and increase the productive life of swards are being oriented
to develop new cultivars with improved disease resistance, looking for germplasm that
shows a stronger root-crown system like the recently released INIA Draco (Altier,
1 997). Inadequate management contributes to limited productive potential and
successful adoption by farmers. There has been a continuous programme of work to
provide management recommendations for new cultivars, often in environments with a
certain degree of restriction.
BFT is regarded as an alternative legume species for New Zealand. However, there has
been renewed interest in the species during the last decade, based on the influence of
condensed tannins on its feeding value (John and Lancashire, 1 98 1 ). However, the most
revolutionary change over traditional cultivars and conceptual models of BFT
morphology and physiology is the recent development in the USA of a prostrate cultivar
with the presence of rhizomes (Beuselinck and Steiner, 1 995). Vegetative reproduction
appears as an alternative focus to solve problems of diseases incidence and plant
reproduction.
Although there has been extensive study of the agronomy of BFT (Smith, 1 962; Smith
and Nelson, 1 967; Nelson and Smith, 1 968; Greub and Wedin, 1 97 1 ; Alison and
Hoveland, 1 989; Li, 1 989), there is little detailed information on the physiological
ecology of the species. Alternative approaches need to be developed to refine BFT
management. Understanding physiological and morphological characteristics of BFT
plants under defoliation, plant architecture to tolerate grazing, and the manipulation of
reproductive processes with a focus on population dynamics, are seen as central issues
to increase BFT persistence.
Chapter One
1.2 OBJECTIVES
- ----------------------- --- - - -
4
Recognising the importance of Lotus corniculatus in pastoral systems as an alternative
legume, and acknowledging the problems in terms of achieving consistency in
production, the main aim of this research project is: to define the best grazing
management to achieve adequate production and persistence in reference to New
Zealand and Uruguayan conditions.
Therefore, the general objectives of this study are as follows:
i) determine appropriate defoliation strategies in terms of frequency, intensity and
timing of defoliation around the year for different Lotus corniculatus cultivars.
ii) quantifY physiological and morphological adaptations under contrasting
management practices.
i ii) analyse population dynamics and strategies to Improve Lotus corniculatus
persistence under grazing conditions.
Additionally, a series of specific objectives relevant to the use of BFT in the conditions
of Uruguay are explored:
i) quantifY the nutritive value of BFT herbage for different cultivars pure or mixed
and under different management strategies.
ii) determine levels of condensed tannins of cultivars under use.
iii) study seed production of BFT under different grazing strategies, and BFT
seedling dynamics to improve plant persistence.
- - ------------------- - - - --
Chapter One
1 .3 THESIS OUTLINE
5
Following a preliminary review of l iterature, experimental results are presented in five
chapters from studies carried out in New Zealand (chapters 3 and 4) and Uruguay
(chapters 5 and 6), and a final section (chapter 7) where a general overview of the
information and the perspectives are presented. A brief description of each chapter
fol lows:
Chapter 2 reviews the relevant l iterature on Lotus corniculatus management, with a
general assessment of its potential as an alternative legume. Focus is on relevant
problems that are limiting Lotus corniculatus utilisation within pastoral systems in New
Zealand and Uruguay.
Chapter 3 focuses on a field experiment oriented to quantify dry matter production,
morphological and physiological changes, and plant survival in Lotus corniculatus, in
response to autumn defoliation and different intensities and frequencies of defoliation in
spring. Winter plant survival and spring regrowth capacity are quantified.
Chapter 4 describes an experiment conducted in glasshouse conditions, oriented to
evaluate the effects of defoliation intensity on biomass allocation during regrowth of
Lotus corniculatus. Regrowth rate, biomass partitioning and morphological changes
under different residual height treatments are evaluated, as well as biomass
accumulation under undisturbed growth conditions. Changes in root carbohydrate
reserves at different growth stages and defoliation intensities are reported.
Chapter 5 presents the results of a field experiment conducted to evaluate the
productive performance of different cultivars of Lotus corniculatus, under contrasting
defoliation regimes during the year of establishment and subsequent effects during the
second year. At the same time, physiological and morphological plant parameters are
described for the cultivars under analysis.
- - - -- ----------------- -- -- -- -
Chapter One 6
Chapter 6 reports a field experiment using a mixed Lotus corniculatus - Trifolium
repens sward to evaluate forage production and BFT stand persistence for different
management systems under grazing conditions. In addition, it relates to seed
production, soil seed bank and seedling emergence of the Lotus corniculatus -
Trifolium repens mixture, providing information on population dynamics in different
grazing systems.
Chapter 7 contains an integrated discussion of the mam findings and general
conclusions, synthesising the information presented in previous chapters and discussing
the essential issues in terms of management recommendations to achieve productive
and persistent Lotus corniculatus pastures. Perspectives for future research are outlined,
emphasising areas that need further detailed study.
1 .3 REFERENCES
ALISON, M.W. AND HOVELAND C.S. ( 1 989). Birdsfoot trefoil management. n. Yield, quality and stand evaluation. Agronomy Journal 81(5) : 745-749.
AL TIER, N. ( 1 997). Enfermedades del Lotus en Uruguay. (Lotus diseases in Uruguay).
INIA La Estanzuela, Uruguay. Serie Tecnica No. 93. ISBN: 9974-38-083-9. 1 6
pg.
BEUSELINCK, P.R. AND McGRA W, R.L. ( 1 989). Environmental considerations for
Lotus production: seed versus herbage. Proceedings of the XVI International
Grassland Congress. Vol. 1 . Nice, France. pp. 649-650.
BEUSELINCK, P.R. AND STEINER, J.1. ( 1 995). Registration of ARS-2620 birdsfoot
trefoil . Electronic address: www .nal .usda.gov Ittic Itektran Idatal 0000061 33
100000633-74.html
BLUMENTHAL, M.1. , KELMAN, W.M., LOLICATO, S . , HARE, M.D. AND
BOWMAN, A.M. ( 1 994). Agronomy and Improvement of Lotus. A review.
"Alternative Pasture Legumes 1993 ". Edited by D.L. Michalk, A.D. Craig and
W.1. Collins. Department of Primary Industries South Australia, Technical
Report 2 1 9. pp. 74-85.
Chapter One 7
BLUMENTHAL, MJ. AND McGRA W, R.L. ( 1 999) . Lotus adaptation, use and
management. Trefoil: The Science and Technology of Lotus. CSSA Special
Publication Number 28. Ed. by P.R. Beuselinck. American Society of
Agronomy, Inc. Crop Science Society of America, Inc. Madison, Wisconsin,
USA. pp. 97- 1 19 .
FORMOSa, F. ( 1 993). Lotus corniculatus. I . Performance forrajera y caracteristicas
agron6micas asociadas. (Productive performance and agronomic characteristics).
Serie Tecnica No. 37. INIA Uruguay. ISBN: 9974-556-69-4 20 pg.
FRAME, J . ; CHARL TON, IF.L. AND LAIDLA W, A.S. ( 1 998). Birdsfoot Trefoil and
Greater Lotus. Chapter 6. In Temperate Forage Legumes. CAB International
ISBN 0-85-1 99-2 1 4-5 . pp. 245-27 1 .
GRANT, W.F. AND MARTEN, G.C. ( 1 985). B irdsfoot trefoil . Forage: The Science of
Grassland Agriculture. Iowa State University. Iowa, USA. Chapter 1 1 , pp. 98-
1 08 .
GREUB, L.l AND WEDIN, W.F. ( 1 97 1 ). Leaf area, dry-matter accumulation, and
carbohydrate reserve levels of birdsfoot trefoil as influenced by cutting height.
Crop Science 1 1 : 734-738.
JOHN, A. AND LANCASHIRE, lA. ( 1 98 1 ). Aspects of the feeding and nutritive value
of Lotus species. Proceedings of the New Zealand Grassland Association 42:
1 52- 1 59.
JONES, D.A. AND TURKINGTON, R. ( 1 986). Biological Flora of the British Isles.
Lotus corniculatus L. Journal of Ecology 74: 1 1 85 - 1 2 12 .
KIRKBRIDE Jr. , lH. ( 1 999). Lotus systematics and distribution. Trefoil: The Science
and Technology of Lotus. CSSA Special Publication Number 28. Ed. by P.R.
Beuselinck. American Society of Agronomy, Inc. Crop Science Society of
America, Inc. Madison, Wisconsin, USA. pp. 1 -20.
LI, Q. ( 1 989). Seed production in birdsfoot trefoil (Lotus corniculatus L.). Thesis for the
Degree of Doctor of Philosophy. Massey University. Palmerston North, New
Zealand. 1 75 P MARTEN, G.C. AND JORDAN, R.M. ( 1 979). Substitution value of birdsfoot trefoil
for alfalfa-grass in pasture systems. Agronomy Journal 71 : 5 5-59.
- - --------------- - - - -
Chapter One 8
McDONALD, H.A. ( 1 946) Birdsfoot trefoil (Lotus corniculatus L.). Its characteristics
and potentialities as a forage legume. Cornell University Agricultural
Experimental Station. Mem. 26 1 . Ithaca, New York.
NELSON, C .J . ; AND SMITH, D. ( 1 968). Growth of birdsfoot trefoil and alfalfa. II.
Morphological development and dry matter distribution. Crop Science 8: 2 1 -25.
PIERRE, J .J . AND JACKOBS, lA. ( 1 953) . The effect of cutting treatments on
birdsfoot trefoil . Agronomy Journal 45: 463-468.
SEANEY, R.R. AND HENSON, P.R. ( 1 970). Birdsfoot trefoil . Advances in Agronomy
22 : 1 1 9- 1 57.
SMITH, D. ( 1 962). Carbohydrate root reserves in alfalfa, red clover, and birdsfoot
trefoil under several management schedules. Crop Science 2: 75-78.
SMITH, D. AND NELSON, C.J. ( 1 967). Growth of birdsfoot trefoil and alfalfa. I .
Responses to height and frequency of cutting. Crop Science 7: 1 30- 1 33 .
STEINER, J.J. ( 1 999). B irdsfoot trefoil origins and germplasm diversity. Trefoil: The
science and technology of Lotus. CSSA Special Publication Number 28. Ed. by
P.R. Beuselinck. American Society of Agronomy, Inc. Crop Science Society of
America, Inc. Madison, Wisconsin, USA. pp. 8 1 -96.
TURKINGTON, R. AND FRANKO, G.D. ( 1 980). The Biology of Canadian Weeds.
4 1 . Lotus corniculatus L. Canadian Journal of Plant Science 60: 965-979.
Chapter Two
2. REVIEW OF THE LITERATURE
2. 1 INTRODUCTION
2.2 AGRONOMIC CHARACTERISTICS
2.2. 1 Plant morphology
2.2.2 Birdsfoot trefoil habitat
2.2.3 Herbage production
2.3 NUTRITIVE VALUE
2.3 . 1 Digestibility
2 .3 .2 Crude protein
2.3 .3 Fibre content
2 .3 .4 Mineral content
2 .3 . 5 Condensed tannins content
2.4 UTILIZATION IN FARMING SYSTEMS
2.4. 1 New Zealand
2.4.2 Uruguay
2.5 DEFOLIA TION MANAGEMENT
2.5 . 1 Frequency and intensity of defoliation and growth habit
2 .5 .2 Leaf area index and photosynthetic efficiency
2 .5 .3 Carbohydrate reserves
2 .6 DISEASE INCIDENCE
2.7 PERSISTENCE REVISITED
2.7. 1 Plant strategy
2.7.2 Improved persistence approach
2 .8 CONCLUSIONS
2.9 REFERENCES
9
Chapter Two
- - - ---------------
2.1 INTRODUCTION
1 0
Birdsfoot trefoil (Lotus corniculatus L.) (BFT) i s a long-lived perennial, tap-rooted and
non-bloating legume, widely distributed and adapted around the world (Seaney and
Henson, 1 970). Many advantages give BFT a recognised popularity for pasture, silage
or hay production in temperate environments. However, it has a reduced production and
poor persistence under intensive grazing (Smith and Nelson, 1 967; Bologna, 1 996).
Additionally, crown-rot diseases have a high incidence, particularly in warm
environments, where poor BFT persistence constitutes an unsolved problem. In fact,
avoidance of hard defoliation regimes and the selection of disease resistant cultivars
played major roles in early development programmes.
In many regions, BFT has limited acceptance and use when compared with other
legumes because of its poor and slow stand establishment. Seed size and seedling
vigour are key l imitations for successful establishment, sometimes overcome by the use
of fertil isers, good inoculation and reducing competition from other species (Twamley,
1 967). During the last decade, BFT has gained popularity due to two main comparative
advantages. Firstly, the interest in low input systems revalidates the place of BFT as an
economic species with low resource requirements. Secondly, BFT feeding value has
particular advantages over many other legume species due to the presence of condensed
tannins (Waghorn et al., 1 998).
In essence, lack of persistence l imits BFT use. An integrated approach to improve the
production and persistence by focusing on the popUlation dynamics of BFT is under
analysis (Olmos, 1 996; Bologna, 1 996; Emery et at. , 1 999). As well, the development
of potential cultivars of BFT with vegetative reproduction (Li and Beuselinck, 1 996;
Beuselinck et al. , 1 996) opened a new window of opportunity for the species.
This chapter will summarise published research on agronomIc and nutritional
characteristics, management requirements and persistence problems currently
Chapter Two 1 1
highlighted for BFT, with particular reference to factors influencing productive
persistence of BFT in New Zealand and Uruguayan pastoral systems.
2.2 AGRONOMIC CHARACTERISTICS
2.2.1 Plant morphology
The main characteristic of the BFT plant is the presence of a deep woody taproot with
numerous lateral branches placed in the first 3 0-60 cm of soil strata, sometimes up to 1
m depth, and accompanied by a dense mass of secondary roots. This structure allows
BFT to perennate like other temperate forage legumes such as lucerne or red clover,
giving advantages in terms of exploration of resources like water and nutrients (Seaney
and Henson, 1 970).
From a crown, a series of primary and secondary shoots are developed. Aerial shoots
are solid and slender, varying from erect to prostrate and having high variation from
glabrous to pubescent (lones and Turkington, 1 986). Branching occurs in the leaf axils
of main and secondary shoots. BFT leaf is pentafoliate (Frame et aI. , 1 998), consisting
of three leaflets attached at the end of the petiole and two others at the base (Seaney and
Henson, 1 970).
The BFT inflorescence is a terminal umbel with 4 to 8 florets, attached by short pedicles
to a long peduncle (Seaney and Henson, 1 970). Pods are long and cylindrical, being
from brown to almost black as maturity advances, and placed at right angles to the top
of the peduncle as a "bird's foot " (Turkington and Franko, 1 980). Seeds are small , with
a 1 000 seed-weight of 1 .2- 1 .4 g (Frame et al. , 1 998). High proportions of hard seeds are
present in mature seed lots, varying from 90% if the crop is hand harvested to 40% if
machine harvested (Brown, 1 955 , cited by Li, 1 989).
Traditionally, two distinct BFT types were recognised: erect European and semi-erect
Empire (Seaney and Henson, 1 970). European types are suited for hay production,
being more erect with faster establishment and regrowth after harvest and earlier
Chapter Two
- - --------------- - - - -
12
maturity than Empire. But they are less tolerant of close grazing, intolerant to poor
drainage and less winter-hardy than Empire types (Seaney and Henson, 1 970;
Turkington and Franco, 1 980). Empire is a semi-erect cultivar, with fine stems and 1 0-
1 4 days later flowering than European types (Grant and Marten, 1 985), and it is
recommended for grazing because of its prostrate habit and indeterminate growth.
However, Empire has slower growth at establishment and subsequently slower regrowth
than European types (Van Keuren and Davis, 1 968).
New Zealand introductions of BFT have been grouped into five categories: South
American, Mediterranean/mild temperate, maritime Europe/cold temperate, continental,
and Middle East (Widdup et al. , 1 987). South American accessions are the most erect
types with low shoot density and moderate frost damage, having poor recovery after
defoliation. Introductions from Mediterranean and mild temperate areas show an
intermediate growth habit, adapted to grazing and forage conservation. They show good
tolerance to drought, but decline under cold and infertile conditions. Continental types
are more prostrate, with high shoot density, well adapted to grazing and resistant to frost
damage. In contrast, the Middle East group is extremely prostrate, but not adapted to
grazing, resembling wild BFT ecotypes.
Wild BFT accessions from Morocco showed a difference from traditional types, in that
they develop rhizomes (Beuselinck et al. , 1 996). In crossing Moroccan material with
traditional American types the rhizomatous character was transferred, developing a
semi-prostrate cultivar with capacity for vegetative reproduction (Beuselinck and
Steiner, 1 995).
2.2.2 Birdsfoot trefoil habitat
BFT grows in Europe, Canada, United States, South America, Australia, New Zealand,
Asia and North Africa (Grant and Marten, 1 985 ; Grant and Small , 1 996), having the
greatest distribution and variation of all species of the genus Lotus (lones and
Turkington, 1 986). Its wide distribution is a consequence of adaptability to a broad
range of soil conditions: acidic, poorly drained, shallow, drought to infertile conditions.
BFT has potential as a replacement for lucerne in less fertile, acid, poorly drained and
Chapter Two
-------------------------------------- - - - -
13
dryland soils where lucerne has limitations (Scott and Charlton, 1 983), and is also
potentially suited for low-input systems (Bullard, 1 990; Charlton and Belgrave, 1 992).
BFT occurs naturally in herb-rich hill and lowland swards, and under adequate
conditions of moisture and fertility and at pH near neutrality grows well (Smith, 1 975).
It is more tolerant to low pH than white clover (Frame, 1 992) or lucerne (Grant and
Marten, 1 985), having a moderate degree of tolerance of drought conditions by virtue of
a deep taproot system (lones and Turkington, 1 986). It is more tolerant to waterlogging
than white clover, red clover and lucerne among others (Heinrichs, 1 970), but
waterlogging causes clorosis, senescence, root decomposition, shoot hypertrophy and
plant death if the period is extended (Vignolio et a!. , 1 994). BFT is less resistant to
waterlogging than Lotus tenuis (Vignolio et aI. , 1 994). It is not well adapted to extreme
high temperatures (Turkington and Franko, 1 980), and performance is poor below 1 2 °C
because of a reduction in symbiotic activity (Kunelius and Clark, 1 970).
Described as a long-day plant, BFT requires 1 6 hours or more day length for full
flowering (Turkington and Franko, 1 980). Under short photoperiod, the rate of
development and number of floral primordia are restricted. The critical photoperiod lies
between 1 4 and 1 4 Yz hours, flowering being very sparse and retarded below this level
(McKee, 1 963). Plant form and height change with photoperiod length, plants being
short, compact, dark green and prostrate under 1 2 hours photoperiod. In contrast, when
photoperiod reaches 1 6 hours, plants are erect, tal l and light green (McKee, 1 963).
Beuselinck and McGraw ( 1 988, 1 989) considered areas below 40° degrees of latitude as
poor environments for BFT seed production in terms of quantity and quality of
production. In these areas, selection of materials with short photoperiod and early
flowering when temperatures are cooler could improve BFT success (Beuselinck and
McGraw, 1 989).
2.2.3 Herbage production
Levels of BFT herbage production can be considered adequate in many regIOns,
because it grows in environments with certain restrictions, where other legumes have a
Chapter Two
---------------- - � - - -- -
14
l imited performance. In Uruguay, BFT forage production reached a maximum during
the second year, with values higher than 9 t DMlha/year (Canimbula et al. , 1 996). When
seed production and seedling recruitment are not allowed, forage production declines in
subsequent years, the best cultivars producing 3.5 t DMlha in the fourth year
(Carambula et al. , 1 996). Growth rate, evaluated in different trials over 1 8 years at La
Estanzuela, Uruguay, showed the lowest range of variation compared with white clover,
red clover and lucerne, achieving average maximum of 42 kg DMIha/day during the
second spring, the maximum registered being 74 kg DMIha/day. In winter, growth rate
of BFT is lower than 1 0 kg DM/ha/day (Diaz et al. , 1 996). Seasonal distribution is
influenced by pasture age, spring-summer production increasing and winter contribution
reducing as pasture age increases (Formoso, 1 993). In Chile, Acufia ( 1 995) reported a
range of 6.7-8.9 t DMlha for the second year, for a group of commercial varieties used
in South America (San Gabriel, Ganador, El Boyero and Quimey), achieving highest
growth in November-December.
Performance of BFT introductions in acid and infertile East Otago soils, New Zealand,
showed good performance from cool temperate and continental types, South American
accessions having the lowest performance for all sites (Fraser et al. , 1 988). Bologna
( 1 996) reported between 7 .5- 1 3 t DM/ha/year in Canterbury, New Zealand, depending
on grazing interval. Growth rate reached a peak of nearly 80 kg DMIha/day in
November-December, declining to winter, when the minimum was less than 1 0 kg
DMIha/day from July to early September
2.3 NUTRITIVE VALUE
BFT forage has a nutritive value comparable to species like lucerne, with the ability to
maintain high quality to advanced stages of maturity. This provides the flexibility to use
BFT in summer when other forage species have low quality, or later as a feed reserve
when grass availability is lower (CoIlins, 1 982; Alison and Hoveland, 1 989 c).
Nutritive value of BFT is associated with season, length of regrowth, or physiological
stage. During vegetative stages, forage contains 60-70% of leaves, declining to maturity
Chapter Two 15
where leaves can contribute only 20-30% (Formoso, 1 993). Grazing management
defines accumulation periods and timing of defoliation, altering the quality of forage
grazed.
2.3.1 Digestibility
BFT has lower organic matter digestibility than white clover (7 1 % vs 85%), due to a
high lignin content in stems compared with other species (John and Lancashire, 1 98 1 ),
suggesting that to achieve an efficient animal performance grazing management should
maximise intake of leaf rather than stem material. Digestibility of leaves remains
constant from early growth to ful l flowering (75%), but stem digestibility declines from
6 1 % during the vegetative stage to 50% at full flowering. After flowering, the decline in
digestibility is high, being associated with a reduction in leaf proportion and an increase
in lignin content of stems (Lopez et al. , 1 965). Early harvests to minimise poor quality
stems and selection of plants with slow decline in stem digestibility with advancing
maturity were proposed by Buxton et at. ( 1 985) .
2.3.2 Crude protein
BFT crude protein (N x 6.25) content ranged from 22.4% during pre-bloom stage to
1 4.5% at seed dehiscence stage (Duell and Gausman, 1 957) . Formoso ( 1 993) identified
a decline in crude protein, from 22% in spring to 1 8% in mid-summer. Crude protein
levels of leaves reported by Scheffer-Basso et al. ( 1 998) varied from 30.3, 26. 1 and
22.9% for vegetative, bloom and full flowering stages, respectively. These authors
reported values of 1 4.8, 1 1 . 1 and 8 .9% of crude protein for stems during vegetative,
bloom and full flowering stages respectively. BFT can retain a high leaf:stem ratio later
than lucerne, indicating the potential for a high intake of crude protein at advanced plant
maturity by animals (Buxton et at. , 1 985 ; McGraw and Marten, 1 986).
Chapter Two
2.3.3 Fibre content
- - ----------- ---- - - -- -
16
Levels of acid detergent fibre reported in a group of BFT materials (El Boyero,
Ganador, San Gabriel and Quimey) showed a range between 24-26% at early vegetative
stage and 3 1 -3 8% at advanced maturity (Acufia, 1 995). Lignin content is lower in leaf
than in stem, increasing with plant maturity because of the high proportion of structural
tissues in stems (Lopez et al. , 1 965).
2.3.4 Mineral content
Nutrient accumulation (mg/kg DM) reaches a maximum earlier than the dry matter
maximum for N, Ca, Mg, S and Zn, only Cu shows a reduced accumulation during this
phase. Accumulation of N, Ca, K, Mg, S, B, Zn and Mn remains stable during seed
fil ling (McGraw et al., 1986). There is a tendency for a reduction in the concentration of
all minerals when maturity advances. Seeds are demanding of N, P, S, Zn and Cu with
concentration two to four times greater than foliar material, suggesting importance in
seed production. The majority of minerals show a greater concentration in leaves than
stems, though in K the reverse is true (Pierre and lackobs, 1 953 ; McGraw et al. , 1 986).
Concentrations of K are higher than 30 g/kg, the maximum tolerable in cattle, but a sole
diet of BFT is unusual (Kallenbach et al. , 1 996). Mg levels ranged up to 4 glkg, which
is the maximum for beef cattle. P concentrations in leaves increase as soil pH increases,
but stems decline, resulting in a deficient diet for animals weighing 450 kg.
Concentrations of Al and Ca are associated with soil pH levels, and for Ca means
exceed critical levels at high pH. Zn in BFT tissues fol lows the same tendency as Zn
concentration in soil, decreasing with pH increase. Mn concentration in tissues
decreases as soil pH increases, but BFT leaves provide adequate levels (30 mg/kg) for
beef cattle (NRC, 1 984, cited by Kallenbach et aI. , 1 996).
Chapter Two 17
2.3.5 Condensed tannins content
BFT and other Lotus specIes are considered non-bloating legumes because of the
presence of condensed tannins (eT) (Kelman and Tanner, 1 990; Ehlke and Legare,
1 993). eT in forage increases protein digestion and utilisation, protecting forage protein
from rumen degradation. Nitrogen retention, protein absorption and essential amino
acids absorption in the small intestine are all increased (John et al. , 1 980; John and
Lancashire, 1 98 1 ; Blumenthal et al., 1 994; Wang et aI., 1 996 c; Min et al. , 1 998).
Despite the benefits described, high eT concentrations can reduce herbage intake,
palatability and digestibility of soluble carbohydrates and hemicel lulose. Kelman and
Tanner ( 1 990) reported that eT concentrations in leaves of accessions of BFT were at
adequate levels to prevent bloating and would not affect dry matter digestibility
negatively.
The presence of eT can substantially improve performance in animals eating BFT. For
example, an increase of 1 4% in wool growth rate (Wang et al., 1 996 c), 1 9% in
efficiency of wool production and improved wool quality (Min et al. , 1 998), higher
rates of carcass gain and higher carcass dressing-out percentage in lambs (Wang et al. ,
1 996 b), a positive effect in milk yield and milk protein of dairy cows (Harris et al.,
1 998) and milk production in lactating ewes (Wang et al. , 1 996 a), have all been
observed in comparisons with animals receiving polyethylene glycol to inactivate eT.
The number of protozoa in rumen fluid of lambs grazing BFT decrease, probably as a
consequence of the astringent nature of eT, or by an inhibition of rumen bacterial
growth and limitations in feed sources for protozoa (Wang et al., 1 996 b). In addition,
the presence of eT in forage has anti-parasitic effects, suggesting advantages in terms
of reduced antihelmintic use (Niezen et al., 1 993 ; Robertson et al. , 1 995).
Additive genetic effects control eT content (Miller and Ehlke, 1 997), which varies
widely among Lotus species. Lotus pedunculatus (OL) shows the highest concentration,
(Kelman and Tanner, 1 990; Roberts and Beuselinck, 1 992), with a range of eT in
leaves from 2.5 to 1 0.7% for a group of 1 0 accessions (Kelman and Tanner, 1 990). BFT
- - - ------------- � - --- -- -
Chapter Two 18
leaves contained between 0. 1 and 7.3% in 22 materials tested, Lotus subbiflorus
moderate to low levels (2 .37-3 .95% in one material) and Lotus tenuis the lowest levels
(0-0.32% in 2 materials tested). eT production is sensitive to environmental conditions, ,
soil fertility and pH. A reduction of 1 0 to 40% in tannin production in high temperature
regimes was reported in the majority of genotypes of BFT tested by Ehlke and Legare
( 1 993), probably by a disruption in normal plant metabolism. Kelman and Tanner
( 1 990) reported no significant lime effects (pH 5 .2), despite a tendency for an increased
eT concentration in unlimed conditions (pH 4.3). Other results showed that soil pH
effects are associated with eT content of genotypes, low tannin types producing less
tannin on acidic soi l , while high and medium types responded inconsistently (Ehlke and
Legare, 1 993).
Recognising the importance of eT in animal production, measurements of eT content
should be incorporated in evaluation and breeding programmes of Lotus species,
particularly in countries where Lotus is widely used, as a way to reduce potential
problems by high eT concentrations. Additional ly, the role of BFT in sward mixtures,
improving nutritive value and nutrient absorption, and acting as a bloating control agent
is under study and should be confirmed in grazing conditions (Waghorn and Shelton,
1 997).
2.4 UTILIZATION IN FARMING SYSTEMS
2.4.1 New Zealand
The earliest description of the use of BFT in New Zealand dates from 1 864 (Thomson,
1 922), and Levy ( 1 9 1 8) first considered its potential agricultural use. However, BFT
remains little used by New Zealand farmers ( 1 000 ha sown in the South Island hill
country, West et al., 1 99 1 cited by Blumenthal and McGraw, 1 999), despite its
recognised productive role. This is probably because of an increasing pressure to
maximise production (Sheath and Hay, 1 989), and a poor understanding of the
agronomy and management principles of BFT. Nodulation failures due to ineffective
inoculation techniques, inability to survive and slow initiation of symbiotic N fixation
Chapter Two 19
are identified problems of BFT establishment in New Zealand environments (Charlton,
1 983 ; Chapman et al. , 1 990; Patrick and Lowther, 1 992). Management
recommendations include inoculation at five times recommended doses, using improved
Rhizobium strains, and dril ling within 24 hours of inoculation, at no more 1 2 mm depth.
Also, it is recommended to sow alone or with non-aggressive grasses. Movement of
rhizobia in the soil for new seedling infection is poor, <0.25 and 4.0 metres/year
lateral ly and downslope respectively, when periods of favourable moisture occur
(Lowther and Patrick, 1 993 ; Douglas et al., 1 996).
The role of BFT in New Zealand pastoral systems depends on its adaptability to
stressful environments, where it competes with many traditionally used legumes (Scott
and Charlton, 1 983). The alternative option is for more fertile and productive soils and
favourable climate, where it is not a good competitor with traditional mixtures of white
clover-ryegrass, that are more productive and persistent under these conditions. In this
case, BFT may be regarded as a complementary feed source in periods where growth or
quality of commonly used mixtures decline, or as a high quality protein bank for special
purposes. The advantages of the presence of condensed tannins reported from many
trials can only be exploited in mixtures if BFT of high persistence and competitive
capacity is used (Waghorn and Shelton, 1 997).
The presence of only one commercially available cultivar of BFT in New Zealand
constitutes a weakness, and exploration of other genetic resources would be useful.
Successful incorporation of alternative species in pastoral systems requires adequate
technical knowledge, as a way to avoid fai lures. Studies to evaluate the effect of
different grazing practices in terms of productivity and long-term persistence will help
to clarify the role ofBFT in New Zealand.
2.4.2 Uruguay
Native grasses and a reduced proportion of legumes are the mam component of
grasslands of Uruguay, growing in acid soils with low phosphorus levels and affected
by irregular drought and wet periods. In this context, the introduction of "pioneer"
- ------------------ - -- --- - -
Chapter Two 20
legumes l ike BFT increases productivity, forage quality and nitrogen inputs. As a
pioneer, it is widely used ( 1 , 1 00,000 ha, including mixtures and pure stands, Altier,
1 995 cited by Blumenthal and McGraw, 1 999), from poor to rich environments and
from extensive to intensive systems.
Considering its priority use for low input systems, the focus for BFT is to achieve
reasonable production and persistence rather than a highly productive performance.
Warm and wet conditions altogether result in a high incidence of diseases, reducing
BFT stands severely in 2-3 years, particularly in areas with a long BFT cultivation
history (Altier, 1 997). Research programmes concentrate on improved persistence by
management practices, or by selecting new materials with improved diseases resistance
(INIA, 1 997). The next step should be to clarify the specific conditions that provide the
best returns from BFT, since currently it is used in a wide range of situations. Often it is
associated with other legumes with different requirements l ike white c lover, resulting in
contradictory management requirements. Under these conditions, productive
performance never achieves high potential, and opportunities for refined management
techniques are l imited.
2.5 DEFOLIATION MANAGEMENT
Several studies have i l lustrated the l imited degree of adaptabi lity of BFT to different
grazing conditions, particularly its weakness to hard grazing. It can not persist at a
productive level under continuous grazing (Van Keuren and Davis, 1 968), and its high
acceptabi lity to l ivestock demands some degree of grazing control (Frame et al. , 1 998).
In fact, BFT performance and tolerance to grazing is strongly influenced by frequency
and severity of defoliation, plant structure and phase of development at defoliation, all
of these factors influencing both short and long-term effects.
Chapter Two
2.5.1 Frequency and intensity of grazing and growth habit
2 1
Pierre and lackobs ( 1 953) reported high forage production during the first year of
production of BFT under close defoliation (2.5 cm). The height of defoliation is less
important under long defoliation intervals, but its effect increases under short intervals
(Smith and Nelson, 1 967). Yield varied by 30% under contrasting defoliation heights,
but by 92% under contrasting defoliation frequency over two years (Smith and Nelson,
1 967).
BFT largely regrows from axillary buds of cut shoots, in contrast to lucerne, which is
more adapted to growth from new buds from the crown. Lucerne responds better to
infrequent cutting, being less tolerant than BFT to hard grazing (Nelson and Smith,
1 968 a). Tall stubble preserves a high proportion of plant parts involved in regrowth,
influencing the vigour of re growth and dry matter yield of BFT (Cordeiro de Araujo and
lacques, 1974). Bologna ( 1 996) found that under two-week defoliation intervals, BFT
re growth was dependent on activation of crown buds. Conversely, with longer intervals
re growth occurred from axillary buds of primary branches.
Prostrate types are less affected than erect types by intense defoliation (Pierre and
lackobs, 1 953 ; Alison and Hoveland, 1 989 c). Changes in defoliation height from 1 0 to
2 .5 cm caused a reduction in plant population of 1 6 and 36% for prostrate and erect
types respectively (Pierre and lackobs, 1 953) . A high residual leaf area on prostrate
materials was identified as the main cause of the difference. Also, Smith and Nelson
( 1 967) described how reduction in winter survival is caused by a low biomass and poor
protection of crown meristems. Autumn harvests also lowered plant stand survival
(Fulkerson, 1 982). During spring-summer, stand losses were caused by diseases (Greub
and Wedin, 1 97 1 a) or poor plant reserves (Alison and Hoveland, 1 989 a).
Despite the drought tolerance described for BFT, careful defoliation during summer is
recommended. Under hard defoliation, plants reduce root mass, losing the capacity to
capture water and nutrients (Vickery, 1 9 8 1 ), and plants prioritise metabolite flux to
aerial rather than underground parts. Root mass declines under intensive defoliation
Chapter Two 22
(Greub and Wedin, 1 97 1 a), and under short defoliation intervals (Smith and Nelson,
1 967; Alison and Hoveland, 1 989 c), the degree of reduction being associated with
plant type (Pierre and Jackobs, 1 953) . In Uruguay, Morales ( 1 992) determined a rapid
decline in BFT proportion by the fourth year of pasture if BFT was defoliated in
summer.
2.5.2 Leaf area index and photosynthetic efficiency
In BFT, maximum crop growth rate was reached at a leaf area index (LAI) of 1 .4 at the
end of spring and remained constant to LAI 6, after flowering, when leaf loss started
(Nelson and Smith, 1 968 b). Growth rate is reduced at the end of summer, leaf losses
starting over LAI 4. BFT growth rate increases over the first 20 days of regrowth, and
maintains its rate during the following 45 days. After that, re growth declines to zero at
80 days approximately (Greub and Wedin, 1 97 1 b). In lucerne, maximum growth was at
a LAI of 3 . 5 , declining before flowering, and loss of leaves started at LAI 6 .5 .
Maximum LAI values in spring were 6. 1 and 6 .7 for BFT and lucerne respectively
(Nelson and Smith, 1 968). In early summer, LAI values between 3-4.5 were reached
after 4 weeks of regrowth if plants were defoliated at 3 . 8 or 1 1 .4 cm height, maximum
being 4 after seven weeks of re growth for all defoliation intensities. In late summer,
LA! was approximately 3 for 7.6 and 1 1 .4 cm height defoliation and 2 for 3 .8 cm height
after seven weeks of re growth. (Greub and Wedin, 1 97 1 a).
The photosynthetic efficiency of re growth declines under long defoliation intervals, old
basal leaves being less efficient. Beuselinck et af. ( 1 984) found that under extended
periods of accumulation, leaves are mainly concentrated in the top canopy strata, with
high leaf losses starting at 60 days. Under long grazing intervals, intense defoliation
could produce a more vigorous regrowth with a greater quantity of new leaves. BFT
residual herbage from grazing after an extended accumulation has poor regrowth
vigour. It i s preferable to use an intensive defoliation to remove old material making a
more efficient use of forage (Twamley, 1 968). Differences in LAI, crop growth rate and
net assimilation rate of BFT and lucerne explain species differences in photosynthetic
efficiency (Gregerson et al., 1 999).
Chapter Two
------------- - - - -
2.5.3 Carbohydrate reserves
23
In BFT, carbohydrate reserves (CHO) are low from spring to autumn even in plants that
grow without defoliation, and these levels are lower than in lucerne or red clover
(Smith, 1 962). After cutting, CHO reserves decline, only increasing again following
photosynthetic activity, explaining the slow regrowth rates and low tolerance to heavy
grazing of BFT (Smith and Soberalske, 1 975). In contrast to the situation with lucerne
or red clover the indeterminate growth habit of BFT, with a constant production of new
shoots, induces a lower CHO accumulation (Smith, 1 962). Greub and Wedin ( 1 97 1 a)
reinforced the description of Smith ( 1 962), both studies showing low levels of CHO
accumulation and its partial effect on BFT re growth, that it is more dependent on C02-
fixation (Gregerson et al. , 1 999). Greub and Wedin ( 1 97 1 a) also proposed that the
metabolic system of BFT differed from that of lucerne, accumulation of CHO requiring
reduction of temperature or photoperiod, conditions achieved in late autumn. Risso et
al. ( 1 983) mentioned that high temperatures reduced photosynthesis and CHO
accumulation in BFT. In red clover, slight growth and high mortal ity rate of plants
growing at high temperatures, was the result of CHO deficiencies (Kendall , 1 958) . In
practice, the reduced patterns of CHO accumulation after BFT defoliation, a difference
from other crown-forming species l ike red clover or lucerne, suggest a high importance
of defoliation height (Nelson and Smith, 1 968 b), the percentages of CHO in roots
increasing with high stubble height fol lowing defoliation (Greub and Wedin, 1 97 1 a).
High levels of root reserves are believed to be essential to improve plant survival under
defoliation or environmental stress. Starch (amylose and amylopectin), sucrose and
reducing sugars (fructose and glucose) are the fractions of total nonstructural
carbohydrate in BFT, being present in similar relative proportions to those in lucerne
(Gregerson et al., 1 999). There is l imited information about seasonal changes and
conversion processes of starch to soluble sugars, as wel l the incidence of other N
containing root compounds on winter survival and regrowth post-defoliation on BFT.
More extensive references are available for lucerne (Fankhauser et al. , 1 989; Li et al. ,
1 996).
Chapter Two
2.6 DISEASE INCIDENCE
24
A diversity of agents, including fungi, pests, root diseases and viruses, are reported as
affecting Lotus spp., in many cases appearing simultaneously or sequentially
(Turkington and Franco, 1 980, Jones and Turkington, 1 986). Stress factors l ike poor
soil drainage, Iow pH, deficiencies in soi l nutrients, adverse winter conditions and
intense defoliation contribute to increased susceptibility of crown forming species to
foIiar and root diseases (Bologna, 1 996). Diseases are a potential l imitation in terms of
productivity and persistence of BFT, drastically reducing stands. In warm and humid
environments losses in stands between 68 and 88% at the end of the second year were
reported as a consequence of disease incidence (Henson, 1 962).
In Uruguay, disease incidence is important after the second spring of the BFT pasture,
total yield in the third year being 25-33% greater if fungicides and insecticides are
applied (Altier, 1 997). The effectiveness of this control i s mainly in stern and leaf
diseases, but it is not effective over crown and root disease complexes which have a
higher impact in BFT persistence. Altier ( 1 997) working with spaced plants, reported
plant losses of 27%, 93% and 99.9% during the first, second and third year respectively.
In Uruguay and Argentina, Fusarium species were the main fungi associated with
crown and root rot diseases, Fusarium oxysporum having the highest relative incidence
(54-58%), in diverse geographic areas (Leath, 1 989). For stern and leaf diseases the
most frequent causal agents are Stemphylium loti and Colletotrichum spp (Chao et al. ,
1 994; Altier, 1 997). Other diseases with minor relevance occur under specific weather
conditions, Sclerotinia trifoliorum being severe under wet and cool weather in fal l and
spring (Altier, 1 994). Also Colletotrichum acutatum produces flower blight when wet
weather occurs during flowering, reducing seed yield in BFT more than 36% (Stewart et
al., 1 994).
In New Zealand, disease research has been concentrated in forage species of agronomic
importance, and alternative species like BFT have received l ittle attention (Watson et
al. , 1 989). Despite that, Scott and Charlton ( 1 983) reported cases of crown and root
diseases (Fusarium species) that severely reduced plant stand. High disease incidence is
Chapter Two 25
reported in warmer and humid areas of the southern United States compared with cold
areas in the north (Grant and Marten, 1 985). Nelson and Smith ( 1 969) found a high
incidence of root diseases when plants grew at 32/24 QC . Under intense defoliation,
plant losses were associated with the incidence of diseases, especially in treatments
defoliated late in summer (Greub and Wedin, 1 9 7 1 a), and plant survival was associated
with high root reserves (Alison and Hoveland, 1 989 a,b). Barta ( 1 978) proposed that
under high temperature, carbohydrate levels are reduced, making plants more easily
affected by pathogen incidence.
Roberts et al. ( 1 994) reported a positive correlation between disease resistance and the
concentration of chitinase, an antifungal hydrolase associated with disease resistance in
other crops. Future disease research will be continuing to develop genotypes with field
resistance, focusing selection on clearly defined disease-defence mechanisms and
understanding ecological population dynamics to develop innovative management
strategies for diseases (English, 1 994). In addition, the effect of pasture management on
the incidence and severity ofBFT diseases could be explored (Altier, 1 994).
2.7 PERSISTENCE REVISITED
From previous sections (2.5 and 2.6), a fragile life cycle was identified in BFT caused
by constraints in management and diseases incidence, BFT being considered in practice
as a short-lived legume under grazing (Pierre and lackobs, 1 953) . To improve BFT
persistence, the understanding of plant strategy to compete, grow and reproduce is
crucial to meet challenges in grazing management oriented to extend productive life of
BFT swards.
2.7.1 Plant strategy
The presence of a deep taproot and a thick fibrous root system in the upper subsoil
provides BFT, like other crown forming species, the capacity to perennate (Forde et al.,
1 989). The structure of the BFT plant allows it to develop and compete in closely
compact growth units or modules, highly branched and slowly expanding, the extended
Chapter Two 26
l ife cycle and reproductive characteristics being classified as "k-strategies" or "phalanx"
species (Pianka, 1 983 ; Hutchings, 1 997). From Grime' s model, BFT can be described
as a competitor species "C", adapted to stable and rich environments (Grime, 1 979),
characterised by low reproductive effort and high growth rate (Grace, 1 99 1 ). As a
competitor, BFT plants have a wel l defined growing season, and capacity to store root
reserves for shoot and root growth, showing a high morphological plasticity for shoot
and root development. (Grime, 1 988). Also, the capacity to grow in low resource
environments determines that BFT can be defined as a stress tolerator species, with
conservative mechanisms of mineral nutrient capture and utilization that constrains
carbon nutrition and reproductive activity (Grime, 1 988). Stress tolerator plants develop
different mechanisms of defence to protect from herbivores and pathogens (Grime,
1 988). Tannins and cyanogenic glycosides are produced by BFT, a costly process for
the plant that is reflected in reduction of growth and reproduction (Briggs, 1 990; Briggs
and Schultz, 1 990).
BFT invests all its assimilates in the maintenance and growth of the individual plant,
traditional BFT types not having the capacity to produce new individuals by vegetative
reproduction. The root/shoot ratio is higher in BFT than in white clover, independent of
environmental conditions, showing the importance of its tap root (Foulds, 1 978). Once
established, BFT survives under drought conditions through its strong root system, a
high root/shoot ratio, low shoot yield and low seed production.
The presence of a rhizomatous characteristic in a new BFT ecotype would allow it to
spread and maintain its population (Beuselinck et al. , 1 996), as also occurs in red clover
(Smith and Bishop, 1 998; Hyslop et a!. , 1 998) and tall fescue (Carlson and Hurst,
1 989). Rhizomatous BFT is expected to avoid crown and root-rot diseases by vegetative
propagation because diseases only affect primary roots. Therefore the plant popUlation
could be maintained without heavy dependence on natural reseeding (Beuselinck et al. ,
1 996).
In Lotus species of contrasting life cycles and morphological structure, plant
investments in belowgroundlaerial mass vary. A rhizomatous plant like Lotus
pedunculatus invests a high proportion of nutrients in the root system and a low
Chapter Two
------------- -- - --
27
proportion in the reproductive system, contrasting with an annual species like Lotus
subbiflorus where resources are assigned mainly to the reproductive process (Olmos,
1 996). BFT appears in an intermediate position; reproductive investments are costly,
but only at the time of peak reproductive output (Briggs and Schultz, 1 990). Lloyd
( 1 980) proposed that in species with indeterminate flowering, the number of flowers is
continually adjusted by environmental conditions. Potential reproductive investments
are regulated by aborting juvenile fruits, altering the number of flowers, altering number
of seeds/fruit, seed mass or intervals between flowering flushes (Stephenson, 1984).
The regulation of reproductive investments determines that seed size or mass is l ittle
altered, an adaptive aspect of high importance in germination, dissemination and
establishment processes (Harper, 1 970, cited by Stephenson, 1 984).
Reported seed production levels on BFT can approach to maximum of 600 kg/ha, but
commonly ranges between 50- 1 75 kg/ha in USA environments (McGraw et al. , 1 986).
2.7.2 Improved persistence approach
The BFT life cycle has four developmental stages: seeds, seedlings, mature vegetative
plants and reproductive plants (Figure 2- 1 ). During the establishment year, seedlings
and vegetative plants are the main components, with low reproductive frequency and
high mortality. The post-establishment phase is characterised by a continued growth of
mature vegetative plants, high frequency of reproduction and low mortality (Emery et
al. , 1 999), popUlation growth being dependent on seed production levels.
Each population is composed of a series of development stages, from seeds to adult
plants. Transitional stages represent the portion of a category that change to others in a
certain period of time (Caswell, 1 989, cited by Emery et al., 1 999). Transitional stages
in Figure 2- 1 comprise the portion of seeds that became seedlings (a2 1 ), the portion of
seedlings that became adults (a32), the adults that became reproductive (a43), the
reproductive success of adult plants improving soil seed reserves (a1 4), and also the
portion of each category that remains stable.
Chapter Two
Time
1
Phases
Pasture establishment
Pasture maintenance
.. Recruitment
Development stages
28
Management objectives
Promote flowering-seeding to increase n1 stage.
To increase n2 stage from early autumn. Promotion of safe niches for recruitment, adding fertilizers to increase seedling g rowth.
Avoid late autumn g razing to reduce n3 stage losses during winter.
Favour the proportion of n4 stage at the end of spring to renew n1 stage.
Figure 2-1. Theoretical model for population dynamics of BFT (partially adapted
from Emery et al. , 1999). Developmental stages are seed (nl), seedling (n2), mature
vegetative plant (n3) and mature reproductive plant (n4). The references al to 4
represent transitional stages from stages nt to n4.
Promotion of flowering-seeding processes will increase soil seed reserves and
encourage the recruitment of new individuals to produce an effective and dynamic
replacement of individuals (Figure 2- 1 ). Maintenance of an adequate plant density will
depend on the survival rate of established individuals and the recruitment rate of new
ones (Jones and Carter, 1 989). Young BFT plants often do not reproduce during the first
year, but matrix population models indicate that with adequate seed production and
recruitment, stand density can be maintained (Emery et al. , 1 999) .
Chapter Two 29
A seed bank like BFT is persistent, where a fraction of the dispersed seed population
survives more than one year as a dormant seed (Hutchings, 1 997). Population growth is
l imited because only a portion of the seed population germinates each year, but this
ensures that at least some seeds germinate during favourable conditions. In Australia,
the size of the seed bank is positively associated with latitude and negatively correlated
with mean maximum temperature in January (Blumenthal and Harris, 1 993, cited by
Blumenthal et al. , 1 994). The proportion of hard seeds in BFT seed banks is high, 38-
4 1 % and 45-54% (Bologna, 1 996; Taylor et al. , 1 973), respectively.
Seedlings ofBFT are described as small , poorly competitive and slow growing (Cooper,
1 967), usually slower than red clover or lucerne (Grant and Marten, 1 985), improving
seedling vigour being a breeding objective (McLean and Nowak, 1 997). BFT seedlings
representing a vulnerable stage in the life cycle are highly intolerant of drought
conditions, (Foulds, 1 978).
Genetic improvement of seedling vigour may be l imited, but major objectives are rapid
expansion rate of cotyledons and true leaves, rapid leaf area expansion, improved leaf
area ratio, rapid re growth rate and fast growing root system (Nelson et al. , 1 994). Seed
size appears important as energy storage only during the first 2-3 weeks, other factors
achieving relevance after that (Nelson et al. , 1 994). BFT seedlings initiated growth in
late summer or early autumn, under a declining photoperiod length, a disadvantage in
competition with aggressive grasses, weeds or companion crops (McKee, 1 962).
BFT seedling survival in New Zealand hill and high country is poor. From 1 28
seedlings/m2 in autumn, less than 1 seedIing/m2 contributed to the stand after one year
(Fraser et al. , 1 994). Recruitment patterns have a peak in autumn, declining during
winter and spring. Despite some advantages in terms of temperature, survival of
emerged spring seedlings is low, due to the dry conditions that affect an undeveloped
root system (Bologna, 1 996). Roberts and Boddrell ( 1 985) identified the spring
emergence peak as the most important for Lotus corniculatus, Medicago lupulina,
Melilotus altissima and Trifolium repens, by breakdown dormancy in winter.
Chapter Two 30
Chapman ( 1 987) reported survival rates lower than 1 0% in Trifolium repens, the
rapidly germinable fraction of the soil seed bank having a small effect on short term
persistence (Nie, 1 997). In Uruguay, seedling emergence of BFT in autumn ranged
from 600- 1 300 seedlings/m2, but establishment was only 1 3-90 plants/m2, an intensive
and successive recruitment each year being proposed from predictive models (Olmos,
1 996). Phosphorus ferti lization increases leaf area index, pod numbers and seed mass of
BFT, improving future recruitment (Olmos, 1 996).
In fact, long term success of BFT requires careful attention to stand management, to
extend the productive life of individuals, minimising stress factors that reduce plant
survival to 2-3 years. The manipulation of reproductive processes to develop a seed
bank needs to be accompanied by an accurate promotion of seedling recruitment in
order to improve plant population. The low efficiency of levels of recruitment supposes
successive phases of seeding-recruitment every year, relatively large seed banks, and
management practices to increase the opportunity for seedling establishment.
2.8 CONCLUSIONS
Considerable progress has been made in improving BFT in pastoral systems;
management practices and the strategic role as a high quality feed source are fields
extensively documented in the literature. Despite this, poor persistence remains as an
unsolved problem. Diseases and abusive management practices are recognised as the
main factors reducing plant populations and consequent productivity. Research is
concentrated on alternative procedures to improve persistence by enhanced reproductive
processes and disease resistance (Figure 2-2). The interaction of management and
disease incidence needs further research, quantifying production losses.
In reproductive processes, efforts are concentrated on aspects of natural reseeding as a
way of maintaining popUlation, and grazing management practices to promote natural
reseeding and consequent recruitment of individuals. Low efficiency of recruitment
processes has been identified in some work, requiring a better definition of the role of
� �- -- -------------- - - - -- -
Chapter Two 31
the soil seed bank, a s well a s all processes involved III seedling recruitment and
survival .
Natural reseeding
r--f Reproduction }- Recruitment strategies
Vegetative propagation
rt Persistence t-Root and crown root complex
Diseases 1 F oliar diseases resistance
Chit/nasa activity
Nutritional processes
H Condensed Feed value tannins J Productive responses
Antihelmintic effects
Population dynamics processes
l Management 1 J Strategic role in productive systems
Figure 2-2. Research priorities in birdsfoot trefoil.
Management strategies play a central role, defining seed bank size and seed quality,
controlling appearance of vegetation gaps for successful establishment, and directing
sward competition in developing stages. The development of a new plant type with
presence of rhizomes opens new expectations in terms of increased plant persistence.
The processes highlighted are partial ly understood, requiring information to enhance
accuracy of management practices. Integration of the effects of management decisions
on aspects of production and persistence provides the focus for studies reported in this
thesis.
Chapter Two
2.9 REFERENCES
32
ACuNA, H. ( 1 995). Comparacion de variedades de tres especies del genero Lotus
(Lotus corniculatus L., Lotus uliginosus Cav. y Lotus tenuis Wald et Kit.) en
suelos de aptitud arrocera. (Comparison of three species of Lotus genus (Lotus
corniculatus L., Lotus uliginosus Cav. y Lotus tenuis Wald et Kit.) in soils
adapted for rice production). INIA Chile. Revista Agricultura Tecnica 58( 1 ): 7-
14 . ISSN: 0365-2807.
ALISON, M.W. AND HOVELAND, C.S. ( 1 989 a). Birdsfoot trefoil management. I .
Root growth and carbohydrate storage. Agronomy Journal 81 : 739-745.
ALISON, M.W. AND HOVELAND, C.S. ( 1 989 b). Root and herbage growth response
of birds foot trefoil entries to subsoil acidity. Agronomy Journal 8 1 : 677-680.
ALISON, M.W. AND HOVELAND C.S. ( 1 989 c). Birdsfoot trefoil management. n. Yield, quality and stand evaluation. Agronomy Journal 81(5) : 745-749.
ALTIER, N. ( 1 994). Current status of research on Lotus diseases in Uruguay.
Proceedings of The First International Lotus Symposium. Edited by Paul R.
Beuselinck, USDA-Agricultural Research Service and Craig A. Roberts,
University of Missouri, Columbia, MO. 22-24 March, 1 994. St. Louis, Missouri,
USA. pp. 203-205 .
ALTIER, N. ( 1 997). Enfermedades del Lotus en Uruguay. (Lotus diseases in Uruguay).
INIA La Estanzuela, Uruguay. Serie Tecnica No. 93 . ISBN: 9974-38-083-9. 1 6
pg.
BARTA, A.L. ( 1 978). Effect of root temperature on dry matter distribution,
carbohydrate accumulation, and acetylene reduction activity in alfalfa and
birdsfoot trefoil. Crop Science 18 : 637-640.
BEUSELINCK, P .R. AND McGRA W, R.L. ( 1 989). Environmental considerations for
Lotus production: seed versus herbage. Proceedings of the XVI International
Grassland Congress. Vol. 1 . Nice, France. pp. 649-650.
BEUSELINCK, P.R., PETERS, EJ. AND McGRAW, R.L. ( 1 984). Cultivar and
management effects on stand persistence of birsfoot trefoil. Agronomy Journal
76: 490-492.
Chapter Two
- --� ----------------
33
BEUSELINCK, P.R. AND STEINER, J.J. ( 1 995). Registration of ARS-2620 birdsfoot
trefoil . Electronic address: www.nal .usda.gov Ittic Itektran Idata/ 0000061 33
100000633-74.html
BEUSELINCK, P.R., LI, B. AND STEINER, J.J. ( 1 996). Plant genetic resources.
Rhizomatous Lotus corniculatus L.: 1 . Taxonomic and cytological study. Crop
Science 36: 1 79- 1 85 .
BLUMENTHAL, MJ., KELMAN, W.M., LOLICATO, S . , HARE, M.D. AND
BOWMAN, A.M. ( 1 994). Agronomy and improvement of Lotus. A review.
"Alternative Pasture Legumes 1 993 ". Edited by D.L. Michalk, A.D. Craig and
WJ. Collins. Department of Primary Industries South Australia, Technical
Report 2 1 9. pp. 74-85 .
BLUMENTHAL, M.J . AND McGRA W, R.L. ( 1 999). Lotus adaptation, use and
management. Trefoil: The Science and Technology of Lotus. CSSA Special
Publication Number 28. Ed. by P.R. Beuselinck. American Society of
Agronomy, Inc. Crop Science Society of America, Inc. Madison, Wisconsin,
USA. pp. 97- 1 1 9.
BOLOGNA, J.J. ( 1 996). Studies on strategies for perennial legume persistence in
lowland pastures. Thesis of Master of Agricultural Science. Lincoln University.
New Zealand. 220 p.
BRIGGS, M.A. ( 1 990). Chemical defense production in Lotus corniculatus L. 1. The
effects of nitrogen source on growth, reproduction and defense. Oecologia 83 :
27-3 l .
BRIGGS, M.A. AND SCHUL TZ, lA. ( 1 990). Chemical defense production in Lotus
corniculatus L. . H. Trade-offs among growth, reproduction and defense.
Oecologia 83 : 32-37.
BULLARD, MJ. ( 1 990). The potential of Bird' s-foot Trefoil for UK Agriculture.
Session 3 . Poster 3 . Research Meeting 2. In British Grassland Society. Second
Research Conference. Scottish Agricultural College, Auchincruive, Ayr, UK.
September 1 990.
BUXTON, D.R., HORNSTEIN, lS. , WEDIN, W.F. AND MARTEN, G.C. ( 1 985).
Forage quality in stratified canopies of alfalfa, birdsfoot trefoil, and red clover.
Crop Science 25: 273-279.
Chapter Two
- -- ---------------- -- -
34
CARAMBULA, M. ; AYALA, W. ; CARRIQUIRY, AND BERMUDEZ, R. ( 1 996).
Informe de resultados de la red de evaluaci6n de forrajeras del Cono Sur
(REFCOSUR). (Forage evaluation report). INIA Uruguay.
CARLSON, 1. T. AND HURST, S . M . ( 1 989). Breeding tall fescue for improved
rhizomatous spreading. Proceedings 0/ the XVI International Grassland
Congress, 4- 1 1 October 1 989, Nice, France. Association Francaise pour la
Production Fourragere, Centre National de Recherche Agronomique, Versail les,
France. pp. 343-344.
CHAO, L.; DE BATTISTA, J. AND SANTINAQUE, F. ( 1 994). Incidence of Birdsfoot
Trefoil crown and root rot in West Uruguay and Entre Rios. Proceedings o/The
First International Lotus Symposium. Edited by Paul R. Beuselinck, USDA
Agricultural Research Service and Craig A. Roberts, University of Missouri,
Columbia, MO. 22-24 March, 1 994. St. Louis, Missouri, USA. pp. 206-208.
CHAPMAN, H. M. ( 1 987). Natural re-seeding and Trifolium repens demography in
grazed hill pastures. n. Seedling appearance and survival . Journal of Applied
Ecology 24: 1 03 7- 1 043 .
CHAPMAN, H. M., LOWTHER, W. L. AND TRAINOR, K. D. ( 1 990). Some factors
limiting the success of Lotus corniculatus in hill and high country. Proceedings
o/the New Zealand Grassland Association 51 : 1 47- 1 50.
CHARLTON, J.F.L. ( 1 983). Lotus and other legumes. In Plant Breeding in New
Zealand. Edited by G.S wratt and H.C. Smith. DSIR. ISBN :0409 70 1 3 7 8 . pp.
253-262.
CHARLTON, J.F.L. AND BELGRA VE, B .R. ( 1 992) . The range of pasture species in
New Zealand and their use in different environments. Proceedings of the New
Zealand Grassland Association 54: 99- 1 04.
COLLINS, M. ( 1 982). Yield and quality of birdsfoot trefoil stockpiled for summer
utilization. Agronomy Journal 74: 1 036- 1 04 1 .
COOPER, C.S. ( 1 967). Relative growth of alfalfa and birdsfoot trefoil seedlings under
low light intensity. Crop Science 7 : 1 75 - 1 78.
CORDEIRO DE ARAUJO, J. AND JACQUES, A. ( 1 974). Caracteristicas morfol6gicas
e producao de materia seca de comichao (Lotus corniculatus L.) colhido em
diferentes estadios de crescimento e a duas alturas de corte. (Morphology and
production of dry matter of birdsfoot trefoil (Lotus corniculatus L.) at different
Chapter Two 35
growth stages and under two defoliation heights) . Revista da Sociedade
Brasileira de Zootecnia. VoI . 3(2): 1 38- 1 47.
DIAZ, 1.E., GARCIA, 1 . AND REBUFFO, M. ( 1 996). Crecimiento de Leguminosas en
La Estanzuela. (Growth of legumes in La Estanzuela). Serie Tecnica No. 71 .
INIA la Estanzuela, Uruguay. ISBN: 9974-38-054-5 .
DOUGLAS, G.B . ; WILLS, B .J . ; PRYOR, H.N.; FOOTE, A.G. AND TRAINOR K.D.
( 1 996). Establishment of perennial legume species in drought-prone, North and
South Island sites. Proceedings of the New Zealand Grassland Association 58:
253-257.
DUELL, R.W. AND GAUSMAN, H.W. ( 1 957). The effect of differential cutting on the
yield, persistence, protein and mineral content of birds foot trefoil . Agronomy
Journal 49: 3 1 8-3 1 9.
EHLKE, N .J. AND LEGARE, D.G. ( 1 993). The effects of temperature and soil stresses
on the production of tannins in birdsfoot trefoil (Lotus corniculatus L.). Lotus
Newsletter Vol. 24.
EMERY, K.M., BEUSELINCK, P., AND ENGLISH, 1 .T. ( 1 999). Evaluation of the
population dynamics of the forage legume Lotus corniculatus using matrix
population models. New Phytologist 144: 549-560.
ENGLISH, J .T. ( 1 994). Diseases of Lotus spp. Proceedings of The First International
Lotus Symposium. Edited by Paul R. Beuselinck, USDA-Agricultural Research
Service and Craig A. Roberts, University of Missouri, Columbia, MO. 22-24
March, 1 994. St. Louis, Missouri, USA. pp. 1 83 - 1 86 .
FANKHAUSER Jr., 1 . ; VOLENEC, 1 .1 . AND BROWN, G.A. ( 1 989). Composition and
structure of starch from taproots of contrasting genotypes of Medicago sativa L.
Plant Physiology 90: 1 1 89- 1 1 94.
FORDE, M.B. , HAY, M.J.M. AND BROCK, J.L. ( 1 989) . Development and growth
characteristics of temperate perennial legumes. In Persistence offorage legumes.
Edited by G.C.Marten, A.G. Matches, R.F. Bames, R.W. Brougham, R.J.
Clements and G.W. Sheath. Proceedings of the Trilateral Workshop, Honolulu,
Hawaii . July 1 988 . ISBN: 0-89 1 1 8-098-2. pp. 9 1 - 1 08 .
Chapter Two 36
FORMOSO, F. ( 1 993) . Lotus corniculatus. 1 . Performance forrajera y caracteristicas
agron6micas asociadas. (Productive performance and agronomic characteristics).
Serie Tecnica No. 37. INIA Uruguay. ISBN: 9974-556-69-4 20 pg.
FOULDS, W. ( 1 978). Response to soil moisture supply in three leguminous species.
New Phytologist 80: 535-545 .
FRAME, 1. ( 1 992). Characteristics of grasses and legumes. Chapter 2. In Improved
Grassland Management. Published by Farming Press Books. Ipswick. ISBN 0
85236 246 3 . pp. 1 6- 1 7
FRAME, J . ; CHARLTON, J.F.L. AND LAIDLAW, A.S. ( 1 998). Birdsfoot Trefoil and
Greater Lotus. Chapter 6. In Temperate Forage Legumes. CAB International,
Wal l ingford. ISBN 0-85- 1 99-2 1 4-5. pp. 245-27 l .
FRASER, WJ., KEOGHAN, 1.M. AND HEENAN, RP. ( 1 988). Performance of Lotus
corniculatus cultivars and lines in an altitudinal sequence on the East Otago
Plateau. Proceedings of the Agronomy Society of New Zealand 18 : 53-57.
FRASER, W.J. , OGDEN, S .C. , WOODMAN, R.F. AND LOWTHER, W.L. ( 1 994).
Role of re-seeding and seedling recruitment for sustainable Lotus corniculatus
based pastures in dry hill and high country. Proceedings of the New Zealand
Grassland Association 56: 1 39- 1 42.
FULKERSON, R. S ( 1 982). Fall harvest effects on the yield and persistence of
birdsfoot trefoil (Lotus corniculatus L.). Lotus Newsletter 13: 9- 1 0.
GRACE, J.B. ( 1 99 1 ). A clarification of the debate between Grime and Tilman.
Functional Ecology 5: 583-587.
GRANT, W.F. AND MARTEN, G.C. ( 1 985). Birdsfoot trefoil . Forage: The Science of
Grassland Agriculture. Iowa State University. Iowa, USA. Chapter 1 1 , pp. 98-
1 08
GRANT, W.F. AND SMALL, E. ( 1 996). The ongm of the Lotus corniculatus
(Fabaceae) complex: A synthesis of diverse evidence. Canadian Journal of
Botany 74: 975-989.
GREGERSON, RG., ROBINSON, D.L. AND VANCE, C.P. ( 1 999). Carbon and
nitrogen metabolism in Lotus. Trefoil: The Science and Technology of Lotus.
CSSA Special Publication Number 28. Ed. by P.R Beuselinck. American
Society of Agronomy, Inc. Crop Science Society of America, Inc. Madison,
Wisconsin, USA. pp. 1 67- 1 85 .
Chapter Two 37
GREUB, L.J. AND WEDIN, W.F. ( 1 97 1 a). Leaf area, dry-matter accumulation, and
carbohydrate reserve levels of birdsfoot trefoil as influenced by cutting height.
Crop Science 1 1 : 734-738.
GREUB, L.l . AND WEDIN, W.F. ( 1 97 1 b). Leaf area, dry-matter accumulation, and
carbohydrate reserves of alfalfa and birdsfoot trefoil under a three-cut
management. Crop Science 1 1 : 34 1 -344.
GRIME, lP. ( 1 979). Plant Strategies and Vegetation Processes. Wiley, Chichester.
GRIME, lP. ( 1 988). The C-S-R model of primary plant strategies-origins, implications
and tests. Chapter 1 4. Plant Evolutionary Biology. Edited by Leslie D. Gottlieb
and Subodh K. Jain.Chapman and Hal l . ISBN 0-41 2-29290 -4. pp. 37 1 -393 .
HARRIS, S .L. , CLARK, D.A. AND LABOYRIE, P.l. ( 1 998). Birdsfoot trefoil - an
alternative legume for New Zealand dairy pastures. Proceedings of the New
Zealand Grassland Association 60: 99- 1 03 .
HEINRICHS, D.H. ( 1 970). Flooding tolerance of legumes. Canadian Journal of Plant
Science 50: 435-438 .
HENSON, P.R. ( 1 962). Breeding for resistance to crown and root rots in Birdsfoot
Trefoil (Lotus corniculatus L.). Crop Science 2 : 429-432 .
HUTCHINGS, MJ. ( 1 997). The structure of plant populations. In Plant Ecology.
Edited by Michael l Crawley. Second edition. ISBN: 0-632-03639-7. Blackwell
Science, Oxford. Chapter 1 1 . pp. 325-358.
HYSLOP, M. G., KEMP, P. D. AND HODGSON, J ( 1 998). Preliminary investigations
into the persistence of a vegetatively reproductive red clover cv. Astred in mixed
swards. Proceedings of the Agronomy Society of New Zealand. 28: 1 1 7- 1 22.
INIA ( 1 997). Plan indicativo de mediano plazo. Documento de circulaci6n interna.
(Research programmes. Internal report)
JOHN, A., ULYATT, MJ., JONES, W.T. AND SHELTON, I .D. ( 1980). Factors
influencing nitrogen flow from the rumen. Proceedings of the New Zealand
Society of Animal Production 40: 226.
JOHN, A. AND LANCASHIRE, lA. ( 1 98 1 ). Aspects of the feeding and nutritive value
of Lotus species. Proceedings of the New Zealand Grassland Association 42:
1 52- 1 59 .
JONES, D.A. AND TURKINGTON, R. ( 1 986). Biological Flora of the British Isles.
Lotus corniculatus L. Journal of Ecology 74: 1 1 85- 1 2 1 2.
Chapter Two
- - - --------------------------------
38
JONES, R.M. AND CARTER, E.D. ( 1 989). Demography of pasture legumes. In
Persistence of Forage Legumes. Edited by G.C.Marten, A.G. Matches, R.F.
Barnes, RW. Brougham, R.J. Clements and G.W. Sheath. Proceedings of the
Trilateral Workshop, Honolulu, Hawaii. July 1 988. ISBN: 0-89 1 1 8-098-2. pp.
1 39- 1 58 .
KALLENBACH, RL, McGRA W, RL. AND BEUSELINCK, P.R. ( 1 996). Soil pH
effects on growth and mineral concentration of birdsfoot trefoil. Canadian
Journal of Plant Science 76(2): 263-267 .
KELMAN, W.M. AND TANNER, G.1. ( 1 990). Foliar condensed tannins in Lotus
species growing on limed and unlimed soil s in South-Eastern Australia.
Proceedings of the New Zealand Grassland Association 52: 5 1 -54.
KENDALL, W.A. ( 1 958). The persistence of red clover and carbohydrate concentration
in the roots at various temperatures. Agronomy Journal 50 :657-659.
KUNELIUS, H.T. AND CLARK, K.W. ( 1 970). Influence of root temperature on the
early growth and symbiotic nitrogen fixation of nodulated Lotus corniculatus
plants. Canadian Journal of Plant Science 50: 569-575.
LEATH, K.T. ( 1 989). Diseases and forage stand persistence in the United States. In
Persistence of Forage Legumes. Edited by G.C Marten; A.G Matches; R.F.
Barnes; R.W. Broughman; RJ. Clements; and G.W. Sheath. Second Printing
1 989. ISBN 0-89 1 1 8-098-2. pp. 465-479.
LEVY, E.B. ( 1 9 1 8). The birdsfoot trefoil. New Zealand Journal of Agriculture 17: 347-
35 1 .
LI, Q. ( 1 989). Seed production in birdsfoot trefoil (Lotus corniculatus L.). Thesisfor the
Degree of Doctor of Philosophy. Massey University. Palmerston North, New
Zealand. 1 75 P LI, B. AND BEUSELINCK, P.R. ( 1 996). Plant genetic resources. Rhizomatous Lotus
corniculatus L. :II. Morphology and anatomy of rhizomes. Crop Science 36: 407-
4 1 1 .
LI, R. ; VOLENEC, J.1 . ; JOERN, B.C. AND CUNNINGHAM, S .M. ( 1 996). Seasonal
changes in non structural carbohydrates, protein, and macronutrients in roots of
alfalfa, red clover, sweetclover and birdsfoot trefoil . Crop Science 36: 6 1 7-623 .
-- - - -��--------�----- -- - -
Chapter Two 39
LLOYD, D.G. ( 1 980). Sexual strategies in plants. I . An hypothesis of serial adjustment
of maternal investment during one reproductive session. New Phytologist 86: 69-
80.
LOPEZ, J. , PRESTES, PJ.Q AND MAGALHAES, E. ( 1 965). A curva de crescimento
e a composiyao em carboidratos soluveis, estruturais, lignina e protein a, e a
digestibilidade em cornichao. (Growth and composition in carbohydrates, lignin,
protein and digestibility of birdsfo,ot trefoil). Proceedings of the IX International
Grassland Congress. Sao Paulo, Brazil. Volume 1 , pp. 85 1 -857 .
LOWTHER, W.L. AND PATRICK, H.N. ( 1 993) . Spread of Rhizobium and
Bradyrhizobium in soil . Soil Biology and Biochemistry 25: 607-6 1 2.
McGRA W, RL. AND MARTEN, G.C. ( 1 986). Analysis of primary spring growth of
four pasture legume species. Agronomy Journal 78: 704-7 1 0.
McGRA W, R.L., RUSELLE, M.P. AND GRA VA J. ( 1 986). Accumulation and
distribution of dry mass and nutrients in birdsfoot trefoil . Agronomy Journal 78:
1 24- 1 3 1 .
McKEE, G.W. ( 1 962). Some effects of daylenght on seeding growth and nodulation in
birdsfoot trefoil. Crop Science 2: 3 1 5-3 1 7.
McKEE, G.W. ( 1 963). Influence of daylength on flowering and plant distribution in
birdsfoot trefoil. Crop Science 3: 205-208.
McLEAN, N.L. AND NOWAK, J. ( 1 997). In vitro selection for improved seedling
vigour in birdsfoot trefoil (Lotus corniculatus L.). Canadian Journal of Plant
Science 77: 3 85-39 1 .
MILLER, P.R AND EHLKE NJ. ( 1 997). Inheritance of condensed tannins in birdsfoot
trefoil . Canadian Journal of Plant Science 77: 5 87-593.
MIN, B.R, BARRY, T.N., MCNABB, W.C. AND KEMP, P. ( 1 998). Effect of
condensed tannins on the production of wool and its processing characteristics in
sheep grazing Lotus corniculatus. Australian Journal of Agricultural Research
49: 597-605 .
MORALES, A. ( 1992). Manejo de la defoliaci6n de Lotus corniculatus sembrado en
cobertura sobre una pastura natural. (Defoliation management of oversown
birdsfoot trefoil). Tesis Ing. Agr. Facultad de Agronomia. Montevideo, Uruguay.
7 1 pg.
- � - �----------------- - - - - - -
Chapter Two 40
NELSON, C.J. ; AND SMITH, D. ( 1 968 a). Growth of birdsfoot trefoil and alfalfa. n. Morphological development and dry matter distribution. Crop Science 8: 2 1 -25 .
NELSON, C.J. AND SMITH, D. ( 1 968 b). Growth of birds foot trefoil and alfalfa. Ill.
Changes in carbohydrate reserves and growth analysis under field conditions.
Crop Science 8 : 25-28.
NELSON, c.J. AND SMITH, D. ( 1 969). Growth of birdsfoot trefoil and alfalfa. IV.
Carbohydrate reserve levels and growth analysis under two temperature regimes.
Crop Science 9: 589-59 1 .
NELSON, C.J. , HUR, S .N. AND BEUSELINCK, P.R. ( 1 994). Physiology of seedling
vigour in birdsfoot trefoil. Proceedings of The First International Lotus
Symposium. Edited by Paul R. Beuselinck, USDA-Agricultural Research Service
and Craig A. Roberts, University of Missouri, Columbia, MO. 22-24 March,
1 994. St. Louis, Missouri, USA. pp. 68-7 1 .
NIE, Z. ( 1 997). The influence of pastoral fallow on hil l country pastures, emphasising
white clover growth behaviour. Thesis for the Degree of Doctor of Philosophy.
Massey University. Palmerston North, New Zealand .. 1 85 p.
NIEZEN, J.H., WAGHORN, T.S . , WAGHORN, G.C. AND CHARLES TON, W.A.G.
( 1 993). Internal parasites and lamb production - a role for plants containing
condensed tannins? Proceedings of New Zealand Society of Animal Production
53 : 235-238.
OLMOS, F. ( 1 996). Estrategias de persistencia y productividad de Lotus. (Strategies for
persistence and productivity in Lotus). Tesis de Maestria en Biologia. Opcion
Ecologia. PEDECIBA. Montevideo. Uruguay. 1 02 pg.
PATRICK, H.N. AND LOWTHER, W.L. ( 1 992). Response of Lotus corniculatus to
inoculation and pelleting on a range of Otago tussock grassland environments.
Proceedings of the New Zealand Grassland Association 54: 1 05- 1 09.
PIANKA, E.R. ( 1 983). Evolutionary Ecology, 3rd. Edition. Harper & Row, New York.
PIERRE, J.J. AND JACKOBS, lA. ( 1 953). The effect of cutting treatments on
birdsfoot trefoil. Agronomy Journal 45: 463-468.
RISSO, D.F., COSCIA, P. AND SURRACO, L. ( 1 983). Productividad de un tapiz de
Lotus (Lotus corniculatus L.) bajo tres manejos del pastoreo. (Productivity of
birdsfoot trefoil under three grazing strategies). Investigaciones Agron6micas
No. 4. Ano 4. No. 1 . pp. 50-56.
Chapter Two 41
ROBERTS, C.A. AND BEUSELINCK, P.R. ( 1 992). Condensed tannins in Lotus
species. Lotus Newsletter . Vol. 23.
ROBERTS, c .A., KARR, A.L., MOHAMMADI, M. , MAREK, S.M. AND
BEUSELINCK, P.R. ( 1 994) . Chitinases in Lotus corniculatus. Proceedings of
The First International Lotus Symposium. Edited by Paul R. Beuselinck, USDA�
Agricultural Research Service and Craig A. Roberts, University of Missouri,
Columbia, MO. 22�24 March, 1 994. St. Louis, Missouri, USA. pp. 1 87- 1 90.
ROBERTS, H.A. AND BODDRELL, lE. ( 1 985) . Seed survival and seasonal pattern of
seedling emergence in some Leguminoseae. Annals of Applied Biology 106:
1 25 - 1 32.
ROBERTSON, RA., NIEZEN, J .H. , WAGHORN, G.c. , CHARLESTON, W.A.G.
AND JINLONG, M. ( 1 995). The effect of six herbages on liveweight gain, wool
growth and faecal egg count of parasitised ewe lambs. Proceedings of the New
Zealand Society of Animal Production 55: 1 99-20 1 .
SCHEFFER-BASSO, S .M. , JACQUES, A.V.A. AND CASTRO, S. ( 1 998). Qualidade
de forragem de especies de Adesmia Dc. e Lotus L. . I. Proteina Bruta e
Digestibilidade in vitro da materia organica. (Quality of Adesmia Dc. e Lotus L.
species. 1. Protein and digestibility). Anais de XVII Reuniao do grupo tecnico
em Forrageiras do Cone SuI. Zona Campos. Utilizacao sustentavel e melhoramento de campos naturais do Cone Sui: Desafios para 0 III Milenio.
Octubro, 1 998. pp. 1 1 6.
SCOTT, D. AND CHARLTON, IF.L. ( 1 983). Birdsfoot trefoil (Lotus corniculatus) as
a potential dryland herbage legume in New Zealand. Proceedings of the New
Zealand Grassland Association 44: 98- 1 05 .
SEANEY, R.R. AND HENSON, P.R. ( 1 970). Birdsfoot trefoil . Advances in Agronomy
22: 1 1 9- 1 57 .
SHEATH, G.W. AND HAY, RJ.M. ( 1 989). Overview of legume persistence in New
Zealand. In Persistence of Forage Legumes. Edited by G.C.Marten, A.G.
Matches, R.F. Bames, R.W. Brougham, RJ. Clements and G.W. Sheath.
Proceedings of the Trilateral Workshop, Honolulu, Hawaii. July 1 988. ISBN: 0-
89 1 1 8-098-2. pp. 23-36.
SMITH, D. ( 1 962). Carbohydrate root reserves in alfalfa, red clover, and birdsfoot
trefoil under several management schedules. Crop Science 2: 75-78.
Chapter Two
------------------ - - -
42
SMITH, D. ( 1 975). Forage management in the North 3rd Edition. Kendall Hunt
Publishers Co., Dubuque, Iowa. 237 p
SMITH, D. AND NELSON, C.l. ( 1 967). Growth of birdsfoot trefoil and alfalfa. I .
Responses to height and frequency of cutting. Crop Science 7: 1 30- 1 33 .
SMITH, D. AND SOBERALSKE, R.M. ( 1 975). Comparison of the growth response of
spring and summer plants of alfalfa, red clover and birdsfoot trefoil. Crop
Science 15 : 5 1 9-522.
SMITH, R. S. AND BISHOP, D. J ( 1 998). Register of Australian herbage plant
cultivars: B. Legumes 1 . Clover (b) Trifolium pratense L. (red clover) cv.
Astred. Australian Journal of Experimental Agriculture 38(3) : 3 1 9-324.
STEPHENSON, A.G. ( 1 984). The regulation of maternal investment In an
indeterminate flowering plant (Lotus corniculatus) . Ecology 65( 1 ): 1 1 3- 1 2 1 .
STEWART, S. , FORMOSO, F . AND ALTIER N . ( 1 994). A flower blight on birdsfoot
trefoil caused by Colletotrichum acutatum. Proceedings of The First
International Lotus Symposium. Edited by Paul R. Beuselinck, USDA
Agricultural Research Service and Craig A. Roberts, University of Missouri,
Columbia, MO. 22-24 March, 1 994. St. Louis, Missouri, USA. pp. 2 1 0-2 1 1 .
TAYLOR T.H., TEMPLETON JR. W.C. AND WYLES J.W. ( 1 973). Management
effects on persistence and productivity of birdsfoot trefoil (Lotus corniculatus
L.). Agronomy Journal 65: 646-648.
TURKINGTON, R. AND FRANKO, G.D. ( 1 980). The Biology of Canadian Weeds.
4 1 . Lotus corniculatus L. Canadian Journal of Plant Science 60: 965-979.
THOMSON, H.G.M. ( 1 922). The Naturalisition of Animals and Plants in New Zealand.
Cambridge University Press, UK. 399 p.
TWAMLEY, B.E. ( 1 967). Seed size and seedling vigour in birdsfoot trefoil. Canadian
Journal of Plant Science 47: 603-609.
TW AMLEY, B .E. ( 1 968). A management study of a birdsfoot trefoil strain trial .
Journal of the British Grassland Society 23: 322-325 .
VAN KEUREN, R.W. AND DAVIS, R.R. ( 1 968). Persistence of birdsfoot trefoil,
Lotus corniculatus L. as influenced by plant growth habit and grazing
management. Agronomy Journal 60: 92-95 .
Chapter Two
- - � ----------------� - �- -� - -
43
VICKERY, PJ. ( 1 98 1 ). Pasture growth under grazing. Chapter 4. In World Animal
Science Vol . B l : Grazing Animals. Edited by F.H.W. Morley. Amsterdam,
Elsevier. pp. 55-7 1 . 405 P VIGNOLIO, 0., MACElRA, N. AND FERNANDEZ O. ( 1 994). Efectos del
anegamiento en invierno y verano sobre el crecimiento y la supervivencia de
Lotus tenuis y Lotus corniculatus. (Effects of winter and summer flooding on
growth and survival of Lotus tenuis y Lotus corniculatus). Ecologia A ustral 4:
1 9-28.
WAGHORN, G.C. AND SHELTON, I.D. ( 1 997). Effect of condensed tannins in Lotus
corniculatus on the nutritive value of pasture for sheep. Journal of Agricultural
Science, Cambridge 1 28 : 365-372 .
WAGHORN, G.C., DOUGLAS, G.B. , NIEZEN, lH., McNABB, W.C. AND FOOTE,
A.G. ( 1 998). Forages with condensed tannins - their management and nutritive
value for ruminants. Proceedings of the New Zealand Grassland Association 60:
89-98.
WANG, Y. , DOUGLAS, G.B. , WAGHORN, G.C. , BARRY, T.N. AND FOOTE, A.G.
( 1 996 a). Effect of condensed tannins in Lotus corniculatus upon lactation
performance in ewes. Journal of Agricultural Science, Cambridge 126: 353-
362.
WANG, Y., DOUGLAS, G.B., WAGHORN, G.C., BARRY, T.N. , FOOTE, A.G. AND
PURCHAS, R.W. ( 1 996 a). Effect of condensed tannins upon the performance
of lambs grazing Lotus corniculatus and Lucerne (Medicago saliva). Journal of
Agriculture Science, Cambridge 126: 87-98.
WANG, Y., WAGHORN, G.C., MCNABB, W.C., BARRY, T.N., HEDLEY, MJ.,
AND SHELTON, I .D. ( 1 996 b) . Effect of condensed tannins in Lotus
corniculatus upon the digestion of methionine and cysteine in the small intestine
of sheep. Journal of Agricultural Science, Cambridge 127: 4 1 3 -42 1 .
WATSON, R.N., SKIPP, R.A., AND BARRATT, LP. ( 1 989). Initiatives in pest and
disease control in New Zealand towards improving legume production and
persistence. In Persistence of Forage Legumes. Edited by G.C.Marten, A.G.
Matches, R.F. Barnes, R.W. Brougham, RJ. C lements and G.W. Sheath.
Proceedings of the Trilateral Workshop, Honolulu, Hawaii. July 1 988. ISBN: 0-
89 1 1 8-098-2. 44 1 -464.
----------------- - - - - -
Chapter Two 44
WIDDUP, K.H., KEOGHAN, J.M., RYAN, D.L. AND CHAPMAN, H. ( 1 987).
Breeding Lotus corniculatus for South Island tussock country. Proceedings of
the New Zealand Grassland Association 48: 1 1 9- 1 24.
Chapter Three 45
3. CHANGES IN THE MORPHOLOGY, PRODUCTION
AND POPULATION OF Lotus corniculatus L. cv.
GRASSLANDS GOLDIE IN RESPONSE TO SEASONAL
DEFOLIATION REGIMES ...
3 . 1 ABSTRACT
3 .2 INTRODUCTION
3 .3 MATERIALS AND METHODS
3 .3 . l Measurements
3 .3 .2 Statistical analysis
3 .4 RESULTS
3 .4. 1 C limate conditions during experimental period
3 .4.2 Herbage production
3 .4.3 Plant density
3 .4.4 Plant morphology
3 .4 .5 Carbohydrate root reserves
3 . 5 DISCUSSION
3 .6 CONCLUSIONS
3 . 7 REFERENCES
... A restricted version of this chapter has been submitted to the Proceedings of
Agronomy Society of New Zealand (Ayala et al., 2000)
Chapter Three
3.1 ABSTRACT
46
A field experiment was conducted from April to December 1 997 at Massey University,
Palmerston North, New Zealand to study responses in morphology, production and
population of birdsfoot trefoil (BFT) (Lotus corniculatus L. cv. Grasslands Goldie) to
seasonal defoliation strategies. A factorial design (2x3x2) was applied in a complete
randomised block arrangement with four replicates on a three year old BFT stand with
94 plants/m2. Treatments included two autumn managements (last cut April or June),
and a combination of two defoliation intervals (20, 40 days) and three defoliation
intensities (2, 6, 1 0 cm) during spring (September-December). Herbage mass, sward
height, botanical composition, plant density and plant morphology parameters (primary
and secondary shoots, root diameter, crown and root mass) were recorded. BFT spring
production reached 3000 kg DM/ha with a mean regrowth rate of 34 kg DM/ha/day.
Early autumn rest (final cut April) improved root carbohydrates content in early spring
and increased BFT spring production 1 7%, but did not affect plant density or the main
morphological parameters. Hard defoliation in spring (2 cm) reduced BFT production
( 1 7%) and plant population (2 1 %) compared with the average of other defoliation
intensities evaluated, and reduced root mass, crown mass, primary and total shoots/m2,
and root reserves. In general, height of defoliation had the greatest effects. Defoliation
frequency did not affect forage production and plant density. Although Grasslands
Goldie is a semi-prostrate BFT, intensive spring defoliation greatly reduced
productivity and persistence, and late autumn uti lisation diminished spring production.
Keywords: Lotus corniculatus L., seasonal management, forage production, plant
morphology, persistence.
Chapter Three
3.2 INTRODUCTION
47
Birdsfoot trefoil (Lotus corniculatus L.) (BFT) has received attention as an alternative
legume species for New Zealand pastoral systems, primarily for its adaptability to less
fertile dryland environments of hill and high country (Scott and Charlton, 1 983 ;
Chapman et al. , 1 990; Charlton and Belgrave, 1 992). It has a high nutritive value for
milk (Woodward et al. , 1 999), wool (M in et al. , 1 998), and meat production (Douglas et
al. , 1 999) partly resulting from the presence of condensed tannins (John et al. , 1 980;
John and Lancashire, 1 9 8 1 ). Grazing management strategies for BFT have received
minor attention, and in many cases fol low recommendations for lucerne (Scott and
Charlton, 1 983), despite it being recognised that BFT has a declining production and
lack of persistence under intensive grazing (Bologna, 1 996). The only BFT variety used
in New Zealand is cv. Grasslands Goldie, a semi-prostrate type adapted to grazing
(AgResearch Grasslands, 1 995). Better persistence is reported for semi-prostrate types
than for upright-growing types of BFT because of differences in the exposure of
meristems to grazing (Van Keuren and Davis, 1 968; Beuselinck et al. , 1 9 84).
The objectives of this research were to determine the influence of the timing of
cessation of defoliation in autumn and the frequency and intensity of spring defoliation
on the productivity and persistence of BFT cv. Grasslands Goldie in a preliminary short
term study.
Chapter Three
3.3 MATERIALS AND METHODS
48
The experiment was conducted on a three year old Lotus corniculatus L. cv. Grasslands
Goldie stand, at the Deer Research Unit, Massey University, Palmerston North, New
Zealand (latitude 40023 'S), from April 1 997 to December 1 997. The soil type was a
deep Tokomaru silt loam (Hewitt, 1 992), with pH 5 .7 and high fertility (Olsen P 24
mg/kg). During summer 1 996/ 1 997 and to the end of March 1 997, the sward was grazed
rotationally at intervals of 30-40 days, at moderate intensities with deer.
The treatments were applied in a 2x3x2 factorial combination, using a randomised
complete block design with four replicates in plots of 3 x 6 m. Initially, there were two
autumn managements with plots cut on 25 April (early autumn rest) or 25 April and 1 0
June (late autumn rest) with a residual height of 3 cm. Thereafter, the plots were rested
until early spring, when a combination of three defoliation intensities (2, 6 or 1 0 cm)
and two defoliation frequencies (20 or 40 days) were introduced. The 2 and 6 cm
defoliation treatments started on 1 0 September, and 1 0 cm defoliation started on 30
September, when BFT Goldie achieved the minimum height required for treatments.
The cutting sequence finished on 1 9 December for all treatments, to allow spelling for
seed production.
Cuts to defined heights were made using a rotary lawn mower, removing herbage from
plots. Plots were sprayed on 7 May with Nortron® (Ethofumesate, 1 .4 llha a.i.) to
control white clover, and on 24 June with Preside® (Flumetsulam, 24 glha a.i.) to
control broadleafweeds, applications being successful in both cases.
Chapter Three 49
3.3.1 Measurements
3.3 .1 .1 Herbage production
Pre-harvest herbage mass was measured at each harvest date and post-harvest mass was
recorded on three occasions, by cutting to ground level with an electric shearing hand
piece in two 500 x 200 mm quadrats per plot. Those quadrat areas were marked to avoid
repeat sampling. Samples were washed to remove residues. Botanical composition was
measured for each sample, by separating into BFT and other components (white clover,
grasses and weeds), and then oven-drying at 60 DC for 48 hours. From each sample, 1 0
stems were randomly collected and dissected into leaves and stems, oven-dried and
weighed. Sward height was evaluated in each quadrat, taking four readings at the top of
the undisturbed BFT sward.
Herbage accumulation (Table 3-2) was calculated as 2: (pre-harvest herbage mass(n+ l ) -
post-harvest herbage mass(n» ' Post-harvest herbage mass (kg DMlha, YPDst) when
quadrat cuts were not taken was estimated from sward height (cm, x) using an equation
constructed with measured post-harvest data.
3.3.1 .2 Plant density, size and morphology
In July, September and December, a soil block of 250 x 250 x 250 mm was taken in
each plot, and manually washed to remove soil and litter. Plants were counted,
recording for each plant the number of primary and secondary shoots. Primary shoots
were defined as the main shoots emerging from the crown, and secondary shoots as
originating in the axils of each primary shoot (Plate 3- 1 ). The diameter of the main root
was measured at a section cut 1 0 mm below the level of insertion of primary shoots.
Crown and roots (> 2 mm diameter) were oven-dried at 60 DC for 48 hours for dry
weight. Crown was defined as the portion above the cut for root diameter measurement
to ground level.
Chapter Three 50
Birdsfoot trefoil morphology
"
\
Secondary shoot
Primary shoot • • • • • • • • •
Crown region
Secondary roots
\, \
Plate 3-1. Birdsfoot trefoil plant showing the morphological parameters evaluated.
3.3. 1 .3 Root carbohydrate analysis
Root samples were collected in September 1 997 (4 plants randomly selected from each
autumn treatment) and December 1 997 ( 1 plant/plot in each treatment). Root samples
were washed, separated from crown and the main root, frozen and freeze-dried.
Samples were ground to pass a 1 mm sieve, making one sample per treatment bulked
across replicates. From samples collected in December, only the early autumn (April)
rest group with all combinations of frequency and intensity of defoliation were
analysed, based on the reduced residual effects tested in other treatments at the end of
the evaluation (December, 1 997). Total available carbohydrates were extracted with
perchloric acid and reaction with anthrone (modified from Clegg, 1 956, see Appendix
1 ). The procedure included the extraction of sugars (glucose, fructose and sucrose) first,
and the extraction of starch (amylose and amylopectin) in a second step (Southgate,
1 99 1 ). In general, extraction with acid provides the same treatment contrasts as
extraction using the enzyme method of Weinmann ( 1 947). Carbohydrate percentages
Chapter Three 51
tend to be slightly higher than from the enzyme method, but involve less analytical time
(Smith, 1 962).
3.3.2 Statistical analysis
Data were analysed by SAS GLM procedures (SAS Institute, 1 990), using a factorial
model in a complete randomised block design for herbage accumulation, plant
morphology and population parameters. Morphology and population parameters
generated from sequential sampling were also analyzed using the 'repeated measures'
option of SAS program. Carbohydrates reserves were not statistically analysed.
3.4 RESULTS
3.4.1 Climate conditions during experimental period
In general, rainfall was 1 8% less than the 60-year mean, with a dry winter in which
rainfall was 55% of the average (Table 3- 1 ). In spring, rainfall was 1 1 % less than
average. Soil temperatures were similar to the 60-year mean ( 1 1 . 1 DC), particularly
during the growing season (AgResearch, Palmerston North).
Table 3-1. Monthly rainfall and soil temperature at 1 00 mm during the evaluation
period and the 60-year average (Source: AgResearch, Palmerston North).
Rainfall (mm) Soil Temperature (OC)
1 997 60-year average 1 997 60-year average
April 1 45 8 1 1 2.5 1 3 .2
May 24 89 1 1 .6 1 0. 1
June 60 97 8 . 1 7.7
July 32 89 6.4 6 .7
August 60 89 7.3 7.6
September 79 75 9 . 1 9.9 October 78 88 1 2.4 1 2.5
November 57 78 1 5 .5 1 5. 1
December 1 03 94 1 7.0 1 7 .3
Chapter Three 52
3.4.2 Herbage production
Forage production of BFT was on average 3000±783 kg DMlha during the spring
period ( 1 0 September - 1 9 December). Herbage masses pre-grazing (kg DMlha, ypre)
and post-grazing (kg DMlha, yposD were positively correlated with sward height (cm, x)
by the equations Ypre = 325.6 + 1 20.4 x (P<O.O I , r2 = 0.72, n = 246) and Ypost =- 1 50.3 +
1 39.4 x (P<O.O I , r2=0.77, n =144), respectively.
Herbage accumulation was significantly affected by autumn management (P<O.O 1 ) and
defoliation intensity in spring (P<0.05), but there were no significant effects of
frequency of defoliation or interaction between main factors (Table 3 -2). Early autumn
rest (April) increased spring production by 1 7%, and forage production was also greater
when managed under 6 cm stubble height than at 2 or 1 0 cm.
Table 3-2. The effect of defoliation management on BFT accumulation (kg
DM/ha), growth rate (kg DMlha/day) and contribution to total production (%) during spring.
BFT accumulation BFT growth rate BFT contribution
(kg DMlha) (kg DM/ha/day) (%) Cutting treatments
Autumn defoliation
Early rest 3290 3 7 65 Late rest 2730 3 1 56 SEM (n) 1 34 (24) 1 .6 (24) 1 .6 (24) Significance * * * * * *
Spring frequency
20 days 2830 32 58 40 days 3 1 80 37 64 SEM (n) 1 34 (24) 1 .6 (24) 1 .6 (24) Significance NS NS *
Spring intensity
2 cm 2690 28 49 6 cm 3400 35 62 I D cm 2920 38 7 1 SEM (n) 1 64 ( 1 6) 1 .9 ( 1 6) 2 .0 ( 1 6) Significance * ** **
P<0.05; * *P<O.O 1 ; NS, not significant; SEM, standard error of the mean; (n), number of observations for each treatment mean
Chapter Three 53
Spring growth rate averaged 34 kg DM/ha/day, and was significantly influenced by
autumn management (P<O.O l ) and defoliation intensity (P<O.O I ). An extended
defoliation period during autumn reduced spring growth by 1 7%. Hard defoliation (2
cm) decreased growth by 24% compared with 6 or 1 0 cm defoliation height (Table 3-2).
Despite the tendency for a high regrowth rate for the 1 0 cm treatment, total
accumulation was lower than for the 6 cm height because of the delay to first harvest
and consequent short accumulation period (80 days vs 1 00 days).
BFT contributed a mean of 6 1 % of total dry mass in spring (Table 3-2), but the BFT
proportion was reduced by delayed autumn rest (June), short defoliation frequency (20
day intervals), and hard defoliation (2 cm height). There were no significant
interactions.
A significant interaction for frequency x intensity of defoliation (P<O.O I ) was detected
for the ratio leaf/(leaf+stem). Under 20 day intervals between defoliation, the proportion
of leaves decreased when defoliation intensity decreased from 2 cm to 1 0 cm residual
height, but at 40 day intervals the reduction in the proportion of leaves was higher in 6
cm than the 1 0 cm height treatment (Table 3 -3) . Leaf proportion of plots cut at 2 cm
and 20 day intervals was sign ificantly higher than for the other treatments, while the
three heights managed at 40 day intervals and the 1 0 cm height at 20 day intervals were
not different, having a ratio leaf:stem near 1 : 1 .
Table 3-3. Leaf/(Leaf+Stem) ratio (dry weight) in BFT under different defoliation
frequencies and intensities in spring.
Frequency Intensity Leaf/(Leaf+Stem) SEM Observations
20 days 2 cm 0.69 0 .0 1 7 52
20 days 6 cm 0 .61 0.0 1 7 5 1
20 days 1 0 cm 0.48 0.02 1 36
40 days 2 cm 0.57 0.024 28
40 days 6 cm 0.46 0.025 28
40 days 1 0 cm 0.5 1 0.024 28
SEM, standard error of the mean
Chapter Three 54
3.4.3 Plant density
During winter, stand reduction was 8% but independent of autumn defoliation
management (P<0.85). Spring plant losses increased significantly (P<O.O l ) with
defoliation intensity, with reductions of 2, 1 6 and 32% for the 1 0, 6 and 2 cm heights
respectively, from September to December (Figure 3- 1 , Plate 3-2). A significant
interaction for time x defoliation intensity (P<0.05) was detected. There were no effects
from autumn management, spring defoliation frequency or other interactions between
treatments on the sampling dates evaluated.
1 00
75 I N
-.§ r.n 50 ......
§ 0:: . 2 cm
25 . 6 cm ... l O cm )I( Winter population
0
July September December
Sampling dates
Figure 3-1. Treatment effects on seasonal changes in birdsfoot trefoil plant density.
Vertical bar represents SEM, (n =16).
Chapter Three 55
Plate 3-2. Number of plants of birdsfoot trefoil in December 1997 under three
intensities of defoliation (2, 6 and 10 cm). Samples represent 250 x 250 mm
quadrat.
3.4.4 Plant morphology
In general, morphological components of plants per unit area were not modified by
autumn management or by spring frequency (Table 3-4) and no significant interaction
effects were observed at the end of the trial. However, root mass/m2 was decreased
(P<0.05) by hard (2 cm) defoliation intensity (Table 3-4), largely due to a reduction in
plant density. Crown mass/m2 was reduced by 45% (P<O.O l ), and total belowground
mass/m2 by 4 1 % in plants defoliated to 2 cm compared with the other defoliation
intensities. Hard defoliation (2 cm) reduced primary shoots/m2 by 49% compared with
more lenient defoliation (Table 3-4). Also, secondary shoots were reduced significantly
(P<0.05) under short defoliation intervals (20 days).
Chapter Three 56
Table 3-4. Effects of defoliation treatments on plant morphology parameters/m2
and plant density in December 1997 for birdsfoot trefoil cv. Grasslands Goldie.
Root Crown Primary Secondary Plant
mass mass shoots shoots Density
Cutting treatments (g/m2) (g/m2) (no.lm2) (no.lm2) (no.lm2)
Autumn defoliation
Early rest 40 84 285 66 1 69 Late rest 38 80 283 689 73 SEM (n) 3 .3 (24) 6.0 (24) 25 (24) 49 (24) 4.7 (24) Significance NS NS NS NS NS
Spring frequency
20 days 37 74 262 598 69 40 days 4 1 90 306 752 73 SEM (n) 3 .3 (24) 6.0 (24) 25 (24) 49 (24) 4.7 (24) Significance NS NS NS * NS
Spring intensity
2 cm 29 53 2 1 2 564 58 6 cm 43 97 3 1 2 692 7 1 I O em 45 96 327 769 84 SEM (n) 4. 1 ( 1 6) 7.3 ( 1 6) 3 1 ( 1 6) 59 ( 1 6) 5 .7 ( 1 6) S i�nificanee * ** * NS * *
N S , not significant; * , P<0.05; * * , P<O.Ol ; SEM, standard error of the mean; (n), number o f observations for eaeh treatment mean
Autumn management affected root diameter in July ( 1 1 and 9 mm for April and June
autumn rest, P<0.05, SEM 0.4), but not in early or late spring. Plants defoliated
frequently and severely (each 20 days and at 2 cm height) showed a reduced root
diameter in December, compared with plants defoliated frequently but at 1 0 cm height
(8 mm vs. 9 mm, P<0.05, SEM 0.4). There were no differences in root diameter
between defoliation intensities when defoliated at intervals of 40 days.
Individual crown and belowground masses were affected by spnng defoliation
frequency (P<0.05), and intensity (P<O.O 1 ), with both increasing under 40 day intervals
or cutting at 6 cm height. Number of primary shoots declined over time (P<O.O 1 ), from
6 shoots/plant in July to 4 shoots/plant in December, but was not affected by
management treatments.
---------------------------------------- - �
Chapter Three
1 5
] 1 0 �------t �% "5i � ] 5 o d3 (/) --.- 20 days interval
___ 40 days interval
o �------�----�------�
2 6 1 0
Defoliation intensity (cm)
57
Figure 3-2. Effects of defoliation management on secondary shoots per plant of
BFT in December, under two intervals and three intensities of defoliation. Vertical
bar represents SEM, (n=8).
In December, a significant cutting frequency x intensity of defoliation interaction was
detected for number of secondary shoots/plant (P<0.05), which declined progressively
from 1 0 cm to 2 cm defoliation height in plants defoliated every 20 days, but were
higher at 2 and 6 cm than 1 0 cm in plants defoliated every 40 days (Figure 3-2).
3.4.5 Root carbohydrate reserves
Results of root carbohydrate analysis should be treated with caution because of the lack
of replication, but treatment contrasts were substantial . From exploratory analysis, root
carbohydrate content in early spring (September) was four times greater for BFT plants
that received an early rest in autumn (April) than in plants with a late autumn rest (June)
(Figure 3 -3) . The free sugar fraction was most affected, being 6 times higher per root
for early rest (April).
Chapter Three
400
April
Timing of autumn rest
o Free sugars
.Starch
June
58
Figure 3-3. Carbohydrate reserves in BFT roots in early spring of plants receiving
two contrasting autumn managements (early rest in April or late rest in June).
400
300
1 00
I 0 Free sugars
I • Starch
I I I
I I� I
o l� J 2cm 6 cm 1 0 cm 2cm 6 cm
20 days interval -- -- 40 days interval
Defo liation treatments
I O em
Figure 3-4. Carbohydrate reserves in BFT roots at the end of spring (December
1 997) of plants managed under two intervals and three intensities of defoliation.
Chapter Three 59
At the end of spring, total carbohydrate reserves were 1 9% higher on average for the 40
day than the 20 day defoliation interval, mainly by an increase in the content of free
sugars (glucose, fructose, sucrose) (Figure 3-4). Plants defoliated at 2 cm height showed
the lowest total carbohydrate content, free sugars and starch content, which was 2.7
times lower than the average of 6 and 10 cm height
3.5 DISCUSSION
The most significant result of this trial was the decline observed in plant density of
BFT, from autumn to the end of spring. The intensity of spring defoliation affected
plant survival, which was severely reduced under hard defoliation (Figure 3 - 1 ). Lesser
defoliation intensities resulted in improved plant survival, and for the case of 1 0 cm
defoliation height the initial spring population was effectively maintained. There were
no effects of defoliation frequency or autumn management on plant survival. Winter
losses were independent of autumn management, suggesting additional factors in that
process.
The age of the BFT sward could partially explain plant losses, as a high incidence of
crown and root diseases is reported for stands 2-3 years old in other regions (Altier,
1 997). In this case, plants taken from the experimental area showed a high incidence of
diseases (Rhizoctonia, Fusarium), when transferred to a warm and humid glasshouse
environment (Chapter 4). Diseases can occur early and progress with the age of the
plant (Leath, 1 989), severe incidence being associated with warm and humid conditions
(Greub and Wedin, 1 97 1 , Beuselinck, et al. , 1 984; English, 1 999). In Otago, New
Zealand, Chapman et al. , ( 1 990) reported losses of 50% of plants of a 3 year old BFT
stand in winter-early spring, a complex of Fusarium species being the causal agent.
Intensive defoliation influenced a number of plant characteristics, resulting in a poor
plant survival. Firstly, root diameter was reduced in plants defoliated severely each 20
days, suggesting a decline in carbohydrate reserve levels. Root carbohydrates analyses
must be interpreted cautiously because they were not replicated, but hard defoliated
Chapter Three 60
plants (2 cm height) showed a content only 30% of values at other defoliation heights
(Figure 3 -4). Plants with low reserve levels and under stress are more susceptible to the
effect of root diseases (Barta, 1 978). Secondly, BFT plants defoliated at both 2 cm and
1 0 cm height had smaller crowns than plants defoliated at 6 cm. It can be suggested that
in 1 0 cm height treatment a wide range of plant sizes could survive. However under
intensive defoliation (2 or 6 cm height) only strong plants could survive, but the
intensive defoliation depleted the crown of surviving plants at 2 cm height. In addition,
the number of secondary shoots was reduced in plants defoliated intensively and
frequently (2 cm - 20 days interval).
These findings at the individual plant level associated with the reduction in plant
density resulted in declining root mass, crown mass and the number of primary stems
per unit of area. The consequence of these associated changes was a reduction in spring
forage production (Table 3 -2), when intensive defoliation was applied. Lax defoliation
is recognised to improve the development of shoots and yield (Cordeiro de Araujo and
Jacques, 1974), and 7- 1 0 cm of residual stubble height is generally recommended for
BFT (Smith and Nelson, 1 967). Better stand persistence under intensive grazing (2.5
cm) was observed only when the stand was defoliated three times during the year
(Smith and Nelson, 1 967). In the current study a reduction of 32% in plant density was
accompanied by 24% reduction in spring forage production, for a stand with 86
plants/m2. Results of Bologna ( 1 996) managing a 1 -2 year old stand of 40 plants/m2 of
BFT Grasslands Goldie, showed spring forage production ranged between 3 . 1 -3 .8 t
DM/ha in environments of South Island of New Zealand, having only 1 3% of reduction
in density if a stand was defoliated at 4 cm height every four weeks and no effects on
density when defoliated at 6-8 week intervals. However, reduction in defoliation
interval to 2 weeks resulted in a decline of 65% in plant density in approximately two
years.
In the current trial (Figure 3-5), the total number of shoots/m2 was closely associated
with BFT contribution, achieving approximately 1 1 00 shoots/m2 with a stand density of
84 plants/m2• Primary shoots contributed significantly to yield when population
increased from 58 to 7 1 plants/m2 by a change in defoliation height from 2 to 6 cm.
However, plants tended to maintain the number of primary shoots when defoliation
Chapter Three
- -� -�-------------------
61
height increased from 6 to 10 cm height, and a further increase in the number of
secondary shoots explained the increase in BFT contribution. Similar determinations
were made by Volenec et al. ( 1 987) for a series of lucerne cultivars, working with a
density range between 1 1 - 1 72 plants/m2; shoot density to obtain high dry matter yields
ranged from 900- 1000 shoots/m2 approximately.
Thus, the combined effect of intensity and frequency of defoliation can alter stand
density drastically, and the indications are that in short periods of time defoliation
height has a stronger effect than frequency of defoliation. Plant density required for
adequate forage production appears to be associated with the age of the stand, because it
declines progressively with sward age. The short number of cutting cycles (one season),
l imited the effects of defoliation interval compared with results for long term trials
(Bologna, 1 996).
� 60 '--" �
,9 ..... ;:l ..0 40 ' I: ... � 0 Q r< � 20 o:l
-e- BFT contribution
-+- Secondary shoots
_ Primary shoots
-I:r-Total shoots
Defoliation height 2 cm 6 cm l O cm
o +---�-··-�-I-------·--
58 71 84
Population in December (plants/m2)
800
� Vl ..... 0 0 ..d r:J'J 400
Figure 3-5. Influence of defoliation height on plant population, shoot density and
herbage contribution to total pasture production in BFT cv. Grasslands Goldie in
spring.
Chapter Three 62
Late autumn defoliation reduced herbage accumulation and growth in spnng, and
reduced root reserves. These findings are in agreement with results of Assuero et al.
( 1 990), who found the levels of carbohydrates reserves in late winter explained 48% of
variation in regrowth of BFT during spring. Winter survival was not affected by autumn
defoliation as reported for other crown forming species l ike lucerne (Keoghan, 1 970)
despite the reduction in carbohydrates root reserves. In the current study, it can be
suggested that winter temperatures were not severe enough to produce the kind of
effects reported for more extreme environments. The stress produced by autumn
defoliation, and consequences to root reserves levels and root diseases, could explain
the 8% reduction in plant density detected in this study. Moreover, weakened plants that
survive winter could die in spring if a proper management is not applied, as is the case
for chicory (Li, 1 997), or for red clover under high temperatures (Kendall, 1958) or dry
conditions (Smith, 1 950), where reserves are deficient
BFT swards under intensive defoliation in autumn also could be less dense and
competitive in spring, increasing the proportion of gaps for the establishment of other
species, by a depletion in root reserves for winter active BFT types or by reduced
competition by low growth of dormant types.
3.6 CONCLUSIONS
Recommended management to optimize forage production and plant persistence ofBFT
cv. Grasslands Goldie in spring should contemplate moderate defoliation intensities (6
cm), independently of defoliation intervals if defoliated for short periods. Management
strategies based on high residual herbage mass after cutting and extended intervals
between defoliations during spring will increase plant survival, but will decrease the
number of grazing cycles possible in the growing season and result in lower quality
forage. However, inappropriate management can reduce foot mass, levels of foot
reserves and crown size of BFT plants, resulting in plants less vigorous and swards less
dense and productive.
Chapter Three 63
No evidence was found to suggest that autumn management has any particular influence
on plant losses during winter in the conditions of North Island of New Zealand.
However, early rest in autumn determined an early and high spring regrowth, based on a
plant population with high carbohydrates root reserves.
3.7 REFERENCES
AGRESEARCH GRASSLANDS ( 1 995). The Grasslands range of forage and
conservation plants. AgResearch Grasslands, Palmerston North. New Zealand.
ALTIER, N. ( 1 997). Enfermedades del Lotus en Uruguay. (Lotus diseases in Uruguay).
Serle Tecnica 93 . INIA La Estanzuela, Uruguay. ISBN: 9974-38-083-9. 1 6 P ASSUERO, S .G., ESCUDER, C.J. , ANDRADE, F. , FERNANDEZ, O. AND
FERNANDEZ, H . ( 1 990) . Efecto de la intensidad de pastoreo sobre 10s
carbohidratos solubles en raices y el rebrote de Lotus corniculatus L. (Effect of
grazing intensity on root reserves and regrowth of Lotus corniculatus L.) Revista
Argentina de Produccion Animal 10(6): 443-453 .
BARTA, A.L. ( 1 978). Effect of root temperature on dry matter distribution,
carbohydrate accumulation, and acetylene reduction activity in alfalfa and
birdsfoot trefoil. Crop Science 18 : 637-640.
BEUSELINCK, P.R., PETERS, E.J. AND McGRAW, R.L. ( 1 984). Cultivar and
management effects on stand persistence of birdsfoot trefoil . Agronomy Journal
76: 490-492.
BOLOGNA, J.J. ( 1 996). Studies on strategies for perennial legume persistence in
lowland pastures. Thesis of Master Agricultural Science. Lincoln University,
New Zealand. 220 p
CHAPMAN, H.M., LOWTHER, W.L. AND TRAINOR, K.D. ( 1 990). Some factors
limiting the success of Lotus corniculatus in hil l and high country. Proceedings
of the New Zealand Grassland Association 51 : 1 47- 1 50.
CHARLTON, J.F.L. AND BELGRAVE, B.R. ( 1 992). The range of pasture species in
New Zealand and their use in different environments. Proceedings of the New
Zealand Grassland Association 54: 99- 1 04.
Chapter Three 64
CLEGG, K.M. ( 1 956). The application of the Anthrone reagent to the estimation of
starch in cereals. Journal of the Science of Food and Agriculture 7: 40-44.
CORDEIRO DE ARAUJO, J. AND JACQUES, A.V.A. ( 1 974). Caracteristicas
morfologicas e prodw;ao de materia seca de cornichao (Lotus corniculatus L.)
colhido em diferentes estadios de crescimento e a duas alturas de corte.
(Morphology and production of dry matter of birdsfoot trefoil (Lotus
corniculatus L.) at different growth stages and under two defoliation heights).
Revista da Sociedade Brasileira de Zootecnia, Vol . 3( 2): 1 38- 147.
DOUGLAS, G.B., STIENEZEN, M., WAGHORN, G.C. AND FOOTE, AG. ( 1 999).
Effect of condensed tannins in birdsfoot trefoil (Lotus corniculatus) and sulla
(Hedysarum coronarium) on body weight, carcass fat depth, and wool growth of
lambs in New Zealand. New Zealand Journal of Agricultural Research 42: 55-
64.
ENGLISH, IT. ( 1999). Diseases of Lotus. In Trefoil:The Science and Technology of
Lotus. CS SA Special Publication No. 28. Edited by P.R. Beuselinck. American
Society of Agronomy, Inc .. Crop Science Society of America, Inc. Madison,
USA. ISBN 89 1 1 8-550-X. pp. 1 2 1 - 1 3 1 .
GREUB, L.J. AND WEDIN, W.F. ( 1 97 1 ). Leaf area, dry-matter accumulation, and
carbohydrate reserves of alfalfa and birdsfoot trefoil under a three cut
management. Crop Science 1 1 : 34 1 -344.
HEWITT, A.E. ( 1 992). New Zealand soil classification. DSIR Land Resources
Scientific Report N° 1 9. Lower Hutt, New Zealand.
JOHN, A, ULYATT, M.J., JONES, W.T. AND SHELTON, I.D. ( 1 980). Factors
influencing nitrogen flow from the rumen. Proceedings of the New Zealand
Society of Animal Production 40: 226.
JOHN, A. AND LANCASHIRE, lA ( 1 98 1 ). Aspects of the feeding and nutritive value
of Lotus species. Proceedings of the New Zealand Grassland Association 42:
1 52- 1 59.
KENDALL, W.A. ( 1 958). The persistence of red clover and carbohydrate concentration
in the roots at various temperatures. Agronomy Journal so: 657-659.
KEOGHAN, lM. ( 1 970). The growth of lucerne fol lowing defoliation. Thesis of
Doctor of Philosophy. Lincoln College, New Zealand. 383 p
LEATH, K.T. ( 1 989). Diseases and forage stand persistence in the United States. In
Chapter Three 65
Persistence of Forage Legumes. Edited by G.C Marten; A.G Matches; R.F.
Bames; R.W. Broughman; KJ. C lements; and G.W. Sheath. Second Printing
1 989. ISBN 0-89 1 1 8-098-2. pp. 465-479.
LI, G. ( 1 997). Response of Chicory (Chichorium intibus L.) to defoliation. Thesis of
Doctor of Philosophy in Plant Science. Massey University, New Zealand. 1 97 p
MIN, B .R., BARRY, T.N., MCNABB, W.C. AND KEMP, P.D. ( 1 998). Effect of
condensed tannins on the production of wool and its processing characteristics in
sheep grazing Lotus corniculatus. Australian Journal of Agricultural Research
49: 597-605 .
SAS INSTITUTE INC. ( 1 990). SAS/STAT user 's guide. Version 6 , 4th edition. SAS
Institute Inc., Cary N.C.
SCOTT, D. AND CHARLTON, IF.L. ( 1 983). B irdsfoot trefoil (Lotus corniculatus L.)
as a potential dryland herbage legume in New Zealand. Proceedings of the New
Zealand Grassland Association 44: 98- 1 05 .
SMITH, D. ( 1 950). Seasonal fluctuations of root reserves in red clover (Trifolium
pratense L.). Plant Physiology 25: 702-7 1 0.SMITH, D. ( 1 962). Carbohydrate
root reserves in alfalfa, red clover and birdsfoot trefoil under several
management schedules. Crop Science 2: 75-78. SMITH, D. AND NELSON, C.J. ( 1 967). Growth of birdsfoot trefoil and alfalfa. I.
Responses to height and frequency of cutting. Crop Science 7: 1 30- 1 33 .
SOUTHGATE, D.A.T. ( 1 99 1 ). Determination of Food Carbohydrates. AFRC Institute
of Food Research, Norwich Laboratory, Norwich, UK. Elsevier Appl ied
Science. ISBN 1 -85 1 66-652-4. Second edition. 232 p
VAN KEUREN, KW. AND DAVIS, R.K ( 1 968). Persistence of birdsfoot trefoil,
Lotus corniculatus L. as influenced by plant growth habit and grazing
management. Agronomy Journal 60: 92-95 .
VOLENEC, J.1. , CHERNEY, lH. AND JOHNSON, K.D. ( 1 987). Yield components,
plant morphology, and forage quality of alfalfa as influenced by plant
population. Crop Science 27: 32 1 -326.
WEINMANN, H. ( 1 947). Determination of total available carbohydrates in plants.
Plant Physiology 22: 279-290.
WOODWARD, S .L . AULDIST, MJ., LABOYRIE, PJ. AND JANSEN, E.B.L. ( 1 999).
Effect of Lotus corniculatus and condensed tannins on milk yield and milk
Chapter Three 66
composition of dairy cows. Proceedings of the New Zealand Society of Animal
Production 59: 1 52- 1 55 .
Chapter Four 67
4. EFFECTS OF DEFOLIATION INTENSITY ON
GROWTH, BIOMASS DISTRIBUTION, AND
MORPHOLOGICAL AND PHYSIOLOGICAL CHANGES
OF BIRDSFOOT TREFOIL
CONDITIONS
4 . 1 ABSTRACT
4.2 INTRODUCTION
4.3 MATERIALS AND METHODS
4.3 . 1 Measurements
4.3 .2 Statistical analysis
4.4 RESULTS
4.4. 1 Growth analysis
4.4.2 Biomass production
4.4.3 Carbohydrate root reserves
4.4.4 Dynamics of plant components
IN GLASSHOUSE
4.4.5 Relationships between herbage harvested and plant components
4.4.6 Plant survival
4.5 DISCUSSION
4.6 CONCLUSIONS
4.7 REFERENCES
Chapter Four
4.1 ABSTRACT
68
The effects of defoliation intensity (2, 6, 1 0 cm cutting height and one undefoliated
control) on biomass production and associated morphological and physiological
changes of B irdsfoot trefoil (Lotus corniculatus L. Grasslands Goldie) were explored in
a glasshouse experiment conducted from May to December 1 997 at the Plant Growth
Unit, Massey University, Palmerston North, New Zealand. B iomass production over
1 20 days under defoliation at 20 day intervals was 1 4, 36 and 44% of the undefoliated
control for 2, 6 and 1 0 cm treatments respectively. Relative growth rate increased
initially in plants defoliated to 2 cm, but after the third harvest (60 days) there were no
differences between treatments. Below-ground biomass decreased over time under hard
defoliation (2 cm). It remained unchanged under 6 cm cutting height and increased for
1 0 cm and the undefoIiated control . The effects of defoliation intensity on plant
components increased significantly over time, with secondary root mass, root diameter
and the number of primary shoots being the most affected parameters. Herbage growth
was positively correlated with crown size, primary roots mass and secondary roots
mass, but not with root diameter. Linear correlation with herbage growth after 80 days
of defoliation was r2=0.66, P<O.O l for crown size, r2=0.38, P<0.05 for primary roots
mass and r2=0.67, P<O.O I for secondary roots mass, r2>0.80, P<O.O l for residual leaf
area and r2=0.62, P<O.OI for root carbohydrate reserves. It can be concluded that BFT
plants under hard defoliation (2 cm) at 20 day intervals cannot produce biomass at
adequate rates . This is despite the adjustments in relative growth rate, increase in leaf
area ratio, specific leaf area or number of leaves per plant. Thus, plants reduce below
ground components severely, reducing the resources for regrowth and affecting plant
survival. More lax defoliation regimes (6 or 1 0 cm) increased productivity and plant
survival .
Keywords: Defoliation, biomass allocation, regrowth, plant components, carbohydrate
reserves.
Chapter Four
4.2 INTRODUCTION
69
The high sensitivity of Lotus corniculatus L. (BFT) to defoliation intensity has often
been observed (Smith and Nelson, 1 967; Greub and Wedin, 1 97 1 ; Alison and
Hoveland, 1 989 a). Recovery after defoliation is associated with the amount, type and
age of tissue removed, as well as environmental constraints, which determine resource
supply (Richards, 1 993). The stubble height after defoliation sets the remaining
photosynthetic active leaf mass to manufacture new plant tissues (Silva, 1 968).
Moreover, the plant structure defines the residual herbage mass and the number of
growing sites, suggesting reasons why more erect BFT cultivars are less persistent
under severe defoliation (Frame et al. , 1 998).
The amount of stored root reserves and their role in re growth after defoliation has been
extensively explored since Graber et al. ( 1 927), with the emphasis being on crown
forming species where the taproot constitutes the main storage organ (Smith, 1 962;
Nelson and Smith, 1 968; Cordeiro de Araujo and Jacques, 1 974; Fankhauser and
Volenec, 1 989). In contrast to other crown-forming legumes, BFT has low levels of root
carbohydrates during the active growing season, with the carbohydrate concentration
not being substantially altered after successive defoliations as occurs in red clover or
lucerne. Root reserves are significantly correlated with forage yield in BFT (Alison and
Hoveland, 1 989 b) and in lucerne (Feltner and Massengale, 1 965), but root weight is not
a good indicator of the amount of reserves in lucerne (Feltner and Massengale, 1 965).
Hodgkinson ( 1 968) found that fine roots of lucerne were reduced after defoliation,
limiting resource uptake.
There is little information in New Zealand l iterature on the physiology and morphology
of BFT after defoliation. The more recent studies refer to morphological responses of
the cultivar Grasslands Goldie to different frequencies of defoliation (Bologna, 1 996).
Previous field work in this thesis using BFT Grasslands Goldie (see Chapter 3), showed
that severe defoliation during spring drastically reduced biomass production and plant
Chapter Four 70
density. Associated effects resulted in reductions in root mass and carbohydrate
reserves, crown size and shoot density of BFT plants in swards defoliated at 2 cm
stubble height. Plant adjustments after successive defoliation cycles in above and
below-ground biomass, potential and relative growth rate, and production of new
photosynthetic tissues are understood for some grasses (Oesterheld, 1 992) and chicory
(Li, 1 997), but there is l imited evidence for BFT. Thus, the objectives of this work were
to provide detailed information on the degree of adaptabil ity and plasticity of BFT cv.
Grasslands Goldie plants to different intensities of defoliation, quantifYing
morphological and physiological plant responses.
4.3 MATERIALS AND METHODS
The experiment was conducted from 1 5 May to 30 December 1 997 in a glasshouse of
the Plant Growth Unit, Massey University, Palmerston North, New Zealand (latitude
40°23 ' S). Three-year-old adult plants of Lotus corniculatus cv. Goldie were collected
from an established sward, selected for uniformity, and then transplanted to plastic
grow-bags ( 1 0.8 litre in volume, 250 mm depth) with three plants per pot.
A standard long term medium (9 months release) was utilised, made up of bark and
amended with dolomite (3 kg/m\ agricultural lime (3 kg/m\ iron sulphate (0.5
kg/m\ and Osmocote plus® (NPK 1 6-3.5- 1 0.8, 4 kg/m\ as a fertiliser. Water was
applied twice daily (5 mins each time) to each pot through an automatic irrigation
system, which watered pots to field capacity (Plate 4- 1 b). Temperatures were
maintained between 1 6±3 .2 and 24±4.6 DC (night/day), heating or ventilating when
necessary. The average minimum and maximum temperatures were 1 4.4 and 32.3 QC,
respectively from September to December.
After transplanting, the pots were maintained from May 1 5 to September 2 without
treatments, giving a period of recovery for the development of secondary roots, and all
pots were defoliated at 6 cm height on July 28 and August 1 (Plate 4- 1 a) .
Chapter Four 71
A completely randomised block design with five replicates was used, with treatments
being three defoliation intensities (2, 6 and 1 0 cm height), and a control treatment that
was not defoliated after the initial cut at 6 cm height when evaluation started. Plants
were defoliated every 20 days at the defined intensities, for 1 20 days from September 2
to December 30, 1 997 (Table 4- 1 ) . Design included 5 internal replicates for each of
four destructive harvests. These were conducted on day 0 ( 1 potlblock, see as pre
destructive harvest), and days 40, 80 and 120 ( 1 pot/treatmentlblock, see as destructive
harvests one to three). The 65 pots under evaluation were grouped on independent
tables per block, and pots in each block rotated weekly.
Table 4-1 . Cutting and destructive harvests schedule for BFT pots from September
to December 1997.
Cuts
2 em
6 em
IO em
Control
Destructive harvests
2 em
6 em
1 0 em
Control
2/9
x X X X
X X X X
2 1 /9
X X X
I I / IO
X X X
X X X X
1 1 1 1
X X X
2 1 1 1 1
X X X
X X X X
1 11 1 2
X X X
30/ 1 2
X X X
X X X X
General management included the control of slugs with Mesurol® pellets (20 g/kg of
carbamate) at 10 g/m2• Plants were sprayed to control aphids with Insectigas®
(dichlorvos in liquid carbon dioxide aerosol) on 6 June. Aphids, white fly and thrips
were controlled on 27 June applying 1 g/l of Orthene® 75 (750 g/kg acephate as a
soluble powder) plus l mlllitre of Attack® (25 gll permethrin plus 475 gll pirimiphos
methyl, emulsifiable concentrate). Aphid control was repeated on 1 4 July, 1 7
November, 5 December and 1 9 December with 1 mlllitre of Attack®.
Chapter Four 72
Approximately 50% of the 200 pots initially prepared showed fungal disease symptoms
(Rhizoctonia and Fusarium) in winter before the start of cutting treatments. The
occurrence was randomly distributed and attributed to high temperature and humid
conditions in the glasshouse (Plate 4- 1 c). Severely affected plants were discharged, and
the remaining pots were drenched with 1 mill of fungicide Saprol® ( 1 90 g/l piperidine) to
control root diseases on 1 3 June and 1 7 July. No further incidence of disease was
apparent.
4.3.1 Measurements
At each cutting date, plant height, herbage mass, number and weight of leaves and
stems and leaf area were measured. All mass data were measured on a dry weight basis
from samples oven-dried at 1 00 QC for 24 hours.
At each destructive harvest, plant height and total above and below-ground masses were
measured. The total above-ground mass was collected in two stages: first above the
defoliation height applied (2, 6 or 1 0 cm height); and then the residual to ground level
to estimate residual herbage mass and components for regrowth of remaining pots.
Primary and secondary shoots were counted, in accordance with definitions of section
3 .3 . 1 .2 of Chapter 3 . A subsample of 20 stems was taken at random from each pot, and
dissected into leaves and stems, number and weight of respective fractions and leaf area
recorded. Leaf area per pot was measured using a LI-Cor LI-3 1 00 leaf area meter
(Lambda Instruments Co., Lincoln, NE, USA).
Roots were washed to remove media, with fine roots collected in a series of sieves after
being washed. Below-ground mass was partitioned into primary roots, secondary roots
and crown mass. Primary root mass included the taproot and part of more lignified
structures (> 2 mm), that could be differentiated from fine and new developing root
tissues. Primary root diameter was measured at a section 1 0 mm below the insertion of
primary shoots.
Chapter Four 73
Main root samples from each pot were collected after washing, kept frozen (-20 °C),
then freeze-dried and ground to pass a 1 mm sieve and stored for carbohydrate analysis
(see Appendix 1 and section 3 .3 . 1 .3 of Chapter 3).
4.3.2 Statistical analysis
Data were analysed with the SAS GLM program (SAS Institute, 1 990), usmg a
complete randomised block design comparing 4 treatments (2, 6 and 1 0 cm defoliation
height and an undefoliated control). All data were initially tested for the assumptions of
normality and homogeneity of variance. Data were transformed using natural
logarithms to homogenise variance when assumptions were not valid, usually due to the
large differences in magnitude of the values of the control treatment with respect to the
other treatments. Parameters quantified repeatedly were analysed by a repeated
measures model over time. Allometric relationships such as leaf area ratio (LAR),
specific leaf area (SLA), weight per leaf and number of leaves per gram were calculated
at each destructive harvest and also analysed by 'repeated measures analysis ' . For
morphological parameters measured at destructive harvests, a multivariate analysis of
variance (MANGY A) was done to identify effects of defoliation intensity. The
relationships between plant components (primary roots mass, secondary roots mass,
crown mass, leaf area and root reserves) and herbage growth were tested using linear
regression models, comparing data of destructive harvests and the fol lowing 20 days
herbage growth.
Chapter Four 74
Plate 4-1 . General view of BFT pots at the time of start (a) and during the trial (b)
and disease symptoms on some BFT plants (c).
Chapter Four
4.4 RESULTS
4.4. 1 Growth analysis
75
The accumulation rate and the relative growth rate (RGR) of herbage harvested were
calculated for each 20-day interval. The accumulation rate was expressed as: g DM
harvested/day (Hodgson, 1 979), and the RGR as: g DM harvested per day/ g DM
residual (Hunt, 1 978). The amount of herbage harvested at each 20-day defoliation was
always in the order 1 0 cm > 6 cm > 2 cm height, and the overall treatment contrast was
always significant (P<O.O I ) (Figure 4- 1 ). However, the main contrast was between 1 0
cm and 6 plus 2 cm height at day 20, and between 1 0 cm plus 6 cm and 2 cm at days
1 00 and 1 20 .
I r
I � 0.75 r � I ! I � 0.5 � � I .<:: I � , o ' " I
0.25 '
* *
I **
** * *
* *
T * * I
0-20 20-40 40-&) 60-80 80-100 100-120 PeriOOs of gr<:Mih (days)
Figure 4-1 . Accumulation rate of herbage harvested (DM g/pot/day) of BFT cv.
Grasslands Goldie defoliated at three intensities in controlled conditions. Vertical
bars represent SEM, (n= number of observations for each treatment mean, were
15, 10, 10, 5, 5 and 5 for 0-20, 20-40, 40-60, 60-80, 80-100 and 1 00-120 day intervals
respectively).
Chapter Four 76
The relative growth rate (RGR) increased when cutting intensity increased (P<O.O 1) in
the first growth cycle (Figure 4-2) . After that, only during the third growth cycle were
differences in RGR significant, the RGR for 1 0 cm defoliation height being higher than
the other treatments.
4
3
o
. 2 cm
1?J6cm
o 10 cm
• •
I
��
NS
r
� � � � � � � �
NS
NS
NS
] • • �
1 � � � � � � � � � �
� � � � � � � � � � � � � � � � �
0-20 20-40 40-60 60-80 80-100 100-120 Periods of growth
Figure 4-2. Relative growth rate (DM gig/day) of BFT cv. Grasslands Goldie
defoliated at three intensities in controlled conditions. Vertical bars represent
SEM, (n= number of observations for each treatment mean, were 15, 10, 10, 5, 5
and 5 for 0-20, 20-40, 40-60, 60-80, 80-100 and 100-120 day intervals respectively).
4.4.2 Biomass production
Above-ground and below-ground biomass were analysed at 40, 80 and 1 20 days of
growth when destructive harvests were conducted (Table 4-2). Data were transformed
using natural logarithms, but actual values are also provided. Cumulative above-ground
growth was defined as the herbage harvested over the respective defoliation height at
each 20 days harvest and accumulated over time. Below-ground mass included the total
root and crown biomass measured at each destructive harvest.
Chapter Four
-------------------------------------- - - -
77
Cumulative above-ground growth at 40 days was significantly different between
treatments (P<O.O I ), with greater biomass when residuals were high. The undefoliated
treatment during the first 40 days did not differ significantly from 1 0 cm height. At 80
and 1 20 days there were significant differences (P<O.O I ) between treatments, with the
control greater than 1 0, 6 and 2 cm height. Defoliation heights of 6 and 1 0 cm were not
significantly different in biomass, and the hardest defoliated (2 cm) treatment had the
lowest biomass. Total biomasses of defoliated treatments were 44, 3 6 and 1 4% of
biomass produced by the control for 1 0, 6 and 2 cm height respectively. Over time,
cumulative growth was significant (P<O.O l ) for all treatments in all cases (Table 4-2).
There were no significant changes over time in the above-ground biomass at time of
destructive harvests for the defoliated treatments, but the control increasing
significantly (P<O.O l , Table 4-2) in biomass over time. At the first destructive harvest,
the above-ground biomass was lower in 2 cm height than in the others. From the second
to the third destructive harvest, the differences between treatments increasing in order
of Control> 10 cm>6 cm>2 cm.
The first destructive harvest produced significant differences between treatments in
below-ground biomass (P<O.O l ), with a tendency for decrease when the intensity of
defoliation was increased (Table 4-2). These differences between treatments were
increased during the subsequent destructive harvests (P<O.O l ).
There was a significant interaction time x defoliation treatment (P<O.O 1 ), with the
control significantly increasing below-ground biomass over time. The 1 0 cm height also
increased biomass, but there were no differences between harvest two (80 days) and
three ( 1 20 days). The 6 cm height maintained its biomass over the time, with no
differences between destructive harvests. On the contrary, the 2 cm treatment reduced
the below-ground biomass, particularly after the first destructive harvest (40 days), with
no differences between destructive harvests two (80 days) and three ( 1 20 days).
The ratio of abovelbelow biomass was highest on average over the experimental period
for 6 cm defoliation.
Chapter Four 78
Table 4-2. Cumulative above-ground growth, above-ground biomass at the time of
destructive harvest and below-ground mass of BFT cv. Grasslands Goldie under
different intensities of defoliation.
Cutting height 40 days
Actual Log
values values
80 days
Actual Log values values
120 days
Actual Log
values values
Cumulative above-ground growth (g DM/potl
2 cm 4. 1 1 .3 1 9 20.9 3.030 37.8 3 .6 1 9 6 cm 6.7 1 .879 42 .9 3.734 94.0 4.535 10 cm 1 2.6 2.53 1 54.4 3.990 1 1 5 .3 4.743 Control 1 3 .3 2.556 87.4 4.456 263.8 5 .566
SEM (n) 0. 1 353 (5) 0.0938 (5) 0.0634 (5) Significance * * * * * *
Above-ground biomass at the time of destructive harvest {g DM/pot}
2 cm 5.2 1 .452 9.6 2. 1 35 6.6 1 .864 6 cm 1 2.9 2.529 22.6 3 .090 24.5 3 . 1 92 10 cm 1 7.6 2 .860 3 1 .9 3.458 32.0 3 .465 Control 2 1 .6 3 .062 95.6 4.552 272. 1 5.598
SEM (n) 0. 1 755 (5) 0. 1 625 (5) 0.0595 (5) Significance * * * * * *
Below-ground mass {g DM/pot}
2 cm 9.5 2.234 4.4 1 .448 4.2 1 .432 6 cm 1 2.2 2.497 1 1 .9 2.458 1 1 .7 2.460 10 cm 1 3 .9 2.6 1 9 24.7 3 . 1 9 1 23.5 3 . 1 35 Control 1 9.2 2.950 73 .8 4.289 2 1 3 .8 5 .355
SEM (n) 0.078 1 (5 ) 0. 1 027 (5) 0.0773 (5) Significance * * * * * *
SEMI Significance
(time-treatment)
0.0882 * *
0. 1409 * *
0.0802 * *
* * , P<O.O I ; SEM, standard error of the mean; (n), number of observations for each treatment mean
Chapter Four 79
4.4.3 Carbohydrate root reserves
There were significant treatment differences in the root carbohydrates content (Table 4-
3). The undefoliated control accumulated starch and free sugars over time particularly
after 80 days growth. Carbohydrate reserves in the control were higher than in
defoliated plants. The starch content differed between defoliation intensities, with the
differences increasing over time (Table 4-3).
Table 4-3. Content of starch and free sugars in roots of BFT cv. Grasslands Goldie
under different intensities of defoliation over 120 days (actual values expressed in
mg/plant and log values used for statistical analysis).
Day 0 Day 40 Day 80 Day 120 SEM I
Cutting Significance
time*treatment
height
Actual Log Actual Log Actual Log Actual Log values values values values values values values values
Starch
2 cm 53 3.940 4 1 3.68 1 22 3 .054 6 cm 243±37 N/A 1 0 1 4.604 62 4 . 129 32 3 .462 0.2230 1 0 cm 1 32 4.880 1 00 4.578 82 4.395 • •
Control 1 60 5.066 475 6. 1 33 3036 8.006
SEM (n) 0. 1 255 (3) 0 . 1439 (3) 0.0750 (3) Significance • • • • • •
Free sugars
2 cm 9 2. 1 35 1 2 2.479 1 7 2.826 6 cm 90± 1 9 N/A 1 2 2.458 1 9 2.9 1 1 25 3.202 0. 1 786 1 0 cm 56 3.996 28 3.336 36 3.588 ••
Control 23 3 . 1 0 1 64 4. 1 1 1 2242 7.663
SEM (n) 0.20 1 5 (3) 0. 1 1 72 (3) 0. 1 300 (3) Significance • • • • • •
* *, P<O.O I ; at day 0 values represent the average±sd (n=3) for all treatments; N/A not statistically analysed; SEM, standard error of the mean; (n), number of observations for each treatment mean
Starch and free sugars rapidly declined after initial defoliation, and for defoliated
treatments remained low and were not replenished during the active growth period. At
all stages starch content decreased significantly on defoliated treatments (P<O.O 1 ), with
increasing defoliation intensity. Also, free sugars content differed significantly at all
stages, the control was the only treatment that increased starch reserves. At the first
Chapter Four
----------------------- - -- -
80
harvest, free sugars content of the control was lower than in the 1 0 cm height.
Differences at the final destructive harvest ( 1 20 days) occurred between 1 0 cm and 2
cm defoliation height, the control being significantly higher than all defoliated
treatments (Table 4-3) .
4.4.4 Dynamics of plant components
Dynamics of below-ground and above-ground plant components were analysed over
time during destructive harvests. Because of differences in magnitude of some
components when compared with the control treatment, data were transformed to
homogenise variance using natural logarithm values for statistical analysis.
4.4.4.1 Below-ground components
Analysis of below-ground components comprised crown mass, primary roots,
secondary roots and root diameter. The original values and time trends are shown in
Figures 4-3 to 4-5, fol lowed by Table 4-4 for the statistical analysis over time based on
transformed data.
25 --0- 2 cm
___ 6 cm 20
� -tr- lO cm ...-0 0..
_ Control � 1 5 on
'-' en en ro S 1 0 � 0 ... u
5
0 0 40 80 1 20
Days
Figure 4-3. Crown mass (g DM/pot) of BFT cv. Grasslands Goldie at cutting
heights of 2, 6 and 10 cm and for an un defoliated control over 120 days.
Chapter Four 81
(a) P r imary roots
1 00 � 2 cm ----.... 0 -e- 6 cm 0.. --::E 75 ---ir- 1 0 c m Cl en ___ Con tro l '-" Vl Vl Cd 50 El Vl ..... 0 0 .....
Q 25 Cd .5 ..... Cl...
0
0 40 80 1 20
Day s
(b) S econdary roots
� 1 25 ��
..... I � 2 cm 0 0.. ---e- 6 cm ::E 1 00 Cl ---ir- ] 0 c m en '-'
Vl 75 ___ Con tro l Vl
Cd El Vl ..... 0 50 0 .....
Q Cd
'"0 25 t:: 0 u Q) VJ 0
0 40 80 1 20
D a y s
Figure 4-4. (a) Primary roots mass and (b) secondary roots mass (DM g/pot) of
BFT cv. Grasslands Goldie at cutting heights of 2, 6 and 10 cm and for an
undefoliated control over 120 days.
Chapter Four
20 r I -<>- 2 cm
5
-.- 6 cm
-I::r- l O cm
_ Contro!
o
82
40 80 120
Days
Figure 4-5. Root diameter (mm) of BFT cv. Grasslands Goldie at cutting heights of
2, 6 and 10 cm and for an undefoliated control over 120 days.
At 40 days, there were differences in crown mass (P< 0.0 1 ) between the control and the
three cutting heights (Table 4-4); but no differences in crown mass between the cutting
heights themselves. At 80 days, crown masses of the control and 1 0 cm height were
significantly greater than for 2 cm height (P<O.O I ). The last destructive harvest ( 1 20
days) had differences (P<O.O l ) between all treatments, with a reduction in crown mass
when the intensity of defoliation increased (Table 4-4).
Over time, there was a significant interaction of time x defoliation treatment (P<O.O 1 )
(Table 4-4). The control treatment significantly increased (P<O.O l ) crown mass over
time, especially at the last destructive harvest ( 1 20 days), which differed from
destructive harvests one (40 days) and two (80 days). For 1 0 cm defoliation height
treatment, crown mass increased significantly over time. The 6 cm height treatment
exhibited no differences in crown size over the experiment. The 2 cm treatment had a
depletion in crown size after the first destructive harvest (P<O.O I ), but not between
destructive harvests two (80 days) and three ( 1 20 days).
Chapter Four 83
2 cm 6 cm
Control
Plate 4-2. Above and below-ground biomass of BFT plants defoliated at different
intensities and undefoliated control at 40 days destructive harvest.
Chapter Four 84
Table 4-4. Crown mass, primary and secondary roots mass and root diameter
(expressed in natural log (x+l) values) of BFT cv. Grasslands Goldie managed at
cutting heights of 2, 6 and 10 cm and one undefoliated control treatment, over 120
days.
40 days 80 days 120 days SEM I Significance
!time*defoliation treatment�
Crown mass
2 cm 1 . 1 97 0.675 0.735 6 cm 1 .303 1 .2 1 8 1 .48 1 0. 1 1 03 1 0 cm 1 .270 1 .70 1 2 . 1 88 * * Control 1 .674 2 .052 2.855
SEM (n) 0.078 1 (5) 0 . 1 448 (5) 0 . 1 039 (5) Significance ** ** * *
Primary roots mass
2 cm 1 .386 0.744 0.756 6 cm 1 .660 1 . 1 62 1 .086 0. 1 445 1 0 cm 1 .82 1 1 .700 1 .420 * * Control 1 . 791 2 .037 4.276
SEM (n) 0. 1 287 (5) 0. 1 482 (5) 0 . 1 524 (5) Significance NS ** * *
Secondary roots mass
2 cm 1 .607 1 . 1 43 1 .065 6 cm 1 .803 2 .079 1 .940 0 . 1 1 08 1 0 cm 1 .9 1 6 2.75 1 2 .490 * * Control 2 .337 4. 1 99 4.822
SEM (n) 0 .0973 (5) 0. 1 4 1 3 (5) 0.088 1 (5) Significance ** ** * *
Root diameter
2 cm 2.405 2 .392 2.260 6 cm 2.442 2 .43 1 2.520 0.0609 1 0 cm 2 .450 2 .509 2.680 NS Control 2 .6 1 1 2 .697 2.759
SEM (n) 0.0435 (5) 0.08 1 2 (5) 0.0594 (5) Significance * NS * *
* * P < 0.0 1 ; * P < 0.05; NS, not significant; SEM, standard error o f the mean; (n), number o f observations for each treatment mean
Chapter Four 85
At 40 days there were no significant differences in primary root mass between the
different defoliation treatments. However, at 80 days, there were significant differences
(P<O.O l ), with a tendency for root mass to be decreased by severe defoliation (Table 4-
4). The undefoliated control and 1 0 cm height treatment did not differ, but were
significantly greater than the 6 and 2 cm heights. Also, primary root mass of 6 cm
height treatment was significantly greater than 2 cm height. During the final destructive
harvest ( 1 20 days), there were differences in main root mass (P<O.O l ) between
treatments, the control having the higher mass and for the other treatments root mass
decreased when intensity of defoliation increased. At 1 20 days, root mass of 1 0 cm
treatment was significantly higher than that of the 2 cm treatment.
Over time, the sequence of destructive harvests showed a significant interaction of time
x defoliation treatment (P<O.O l ), the undefoliated control increased primary root mass
after the second destructive harvest (80 days) (Table 4-4). For the 1 0 cm height, the last
destructive harvest ( 1 20 days) showed a reduction in root mass compared with the first
destructive harvest (40 days). For 2 and 6 cm height, there were reductions in main root
mass after first destructive harvest (40 days), but no significant differences between
destructive harvests two (80 days) and three ( 1 20 days).
At all destructive harvests there were significant differences (P<O.O l ) in secondary root
mass between treatments (Table 4-4). At the first destructive harvest (40 days) only the
control had a higher mass of secondary roots than the other treatments. After 80 days,
the control had the highest root mass, more than three times greater than lax 1 0 cm
defoliation treatment (Figure 4-4 b). The mass of secondary roots diminished when
intensity of defoliation increased, behaviour that was confirmed during the last
destructive harvest ( 1 20 days).
There was a significant interaction time x defoliation treatment (P<O.O l ). The control
treatment increased secondary root mass over the time, with significant differences
between the three destructive harvests (Table 4-4). For the 1 0 cm defoliation height,
secondary root mass increased between 40 and 80 days, but there were no differences
between destructive harvests two (40 days) and three (80 days). For the 6 cm treatment,
Chapter Four 86
there were no differences over the time, contrasting with the reduction in secondary
roots for the 2 cm treatment.
During the first and third destructive harvests there were differences in the root
diameter of BFT plants under different defoliation intensities (P<0.05 and P<O.O l ,
respectively). At 40 days, only the control differed from the other treatments, with
diameter 2 mm greater than the best of the defoliated treatments ( 1 0 cm height) (Figure
4-5, Table 4-4). At 1 20 days, the control and 1 0 cm height were not different and had a
greater root diameter than 2 cm height treatment. Also, the control had greater root
diameter than 6 cm height, but 6 cm did not differ from 1 0 cm height (Table 4-4).
Over time, the undefoliated control did not show changes in root diameter (Table 4-4),
whereas the 1 0 cm defoliation height increased root diameter when the first destructive
harvest was compared with the third (P<0.05). If the pre-treatment harvest is
considered, root diameter increased for 1 0 cm and control treatments. The root diameter
of plants defoliated at 6 cm height remained unchanged over 120 days, but those
defoliated at 2 cm height reduced in diameter, particularly at the last destructive harvest
( 1 20 days).
4.4.4.2 Above-ground components
Shoot numbers, leaf/(leaf+stem) ratio and allometric relationships (LAR, SLA,
weight/leaf and leaves/gram) were analysed over time (see Tables 4-5 and 4-6).
The number of primary shoots was not affected at the first destructive harvest (40 days),
but a significant decline (P<O.O l ) under hard defoliation (2 cm) was observed at 80
days (Table 4-5). At 1 20 days control plants had more primary shoots than all defoliated
treatments (P<O.O 1 ) . Between defoliated treatments, only 6 cm had a higher number of
primary shoots than the 2 cm height. Over time, a significant interaction of time x
defoliation treatment was observed (P<O.O I ). The undefoliated control increased
primary shoots at the final destructive harvest ( 1 20 days), but 2 cm and 1 0 cm treatment
reduced shoots per pot and the 6 cm treatment remained unchanged.
Chapter Four 87
The number of secondary shoots was significantly higher in the control (P<0.05, Table
4-5) compared with 2 and 1 0 cm height, but defoliated treatments did not differ at first
destructive harvest (40 days). During the second destructive harvest (80 days), the
control had a higher number of secondary shoots than the 2 and 1 0 cm height, and the
more lenient defoliation treatments (6 and 1 0 cm height) were higher than for the 2 cm
defoliation height. At 1 20 days, the control remained higher than the 2 or 6 cm
treatment. Between defoliated treatments, secondary shoots decreased significantly
when the defoliation intensity increased. Over time, there was a significant interaction
(P<O.O 1 , Table 4-5), treatments decreased in shoot density after the initial increase at 40
days with the exception of 1 0 cm height that maintained shoots density over time.
The leaf/(Ieaf+stem) ratio was not affected by defoliation intensity, but it declined over
time (P<0.05). The average was 0.62, 0.48 and 0.40 from the first to the third
destructive harvests respectively (SEM 0.03 1 ) (data not shown).
The LAR at the first destructive harvest (40 days) showed differences (P<O.O 1) between
treatments, the more lax the defoliation the higher LAR, and the control was not
different from 1 0 or 6 cm height. At day 80, there were no differences between
treatments, but at the final destructive harvest ( 1 20 days) there were differences
(P<0.05). In this case, the control had the lowest LAR and, contrasting with first
destructive harvest, 2 cm height had the highest value (Table 4-6). There was a
significant interaction of time x defoliation treatment (P<0.05, Table 4-6), 2 and 6 cm
increased LAR over time, but 1 0 cm and the control remained unchanged.
Chapter Four 88
Table 4-5. Primary and secondary shoots per pot (actual and log values) of BFT
cv. Grasslands Goldie managed at cutting heights of 2, 6 and 10 cm and one
undefoliated control treatment, over 120 days.
40 days 80 days 120 days SEM I Significance
(time x height) Actual Log Actual Log Actual Log values values values values values values
Primary shoots
2 cm 1 4 2 .62 5 1 .39 8 2 .09 6 cm 1 3 2.57 1 0 2 .23 1 5 2.59 0. 1 42 1 0 cm 1 8 2 .84 1 6 2 .73 1 2 2 .41 ** Control 1 1 2 .35 1 5 2.53 30 3 .34
SEM (n) 0 . 1 1 3 (5) 0.202 (5) 0. 1 52 (5) Significance NS * * * *
Secondary shoots
2 cm 22 2 .9 1 9 2.3 1 9 2.30 6 cm 27 3.25 27 3 .30 1 2 2 .58 0 . 1 48 1 0 cm 1 8 2.88 2 1 3 .09 20 3 .00 ** Control 48 3 .83 35 3 .54 1 7 2.87
SEM (n) 0.2 1 6 (5) 0.09 (5) 0.06 (5) Significance * * * * *
* * P< 0.0 1 ; * P< 0.05; NS, not significant; SEM, standard error of the mean; (n), number of observations for each treatment mean
The SLA was only affected by defoliation height at the first destructive harvest (P<0.05,
Table 4-6). The 10 cm and the control treatments showed higher SLA than 2 cm height
and also 1 0 cm differed from 6 cm height. A significant interaction of time x defoliation
treatment was observed (P<0.05); the defoliated treatments, but not the undefoliated
control, increased in SLA from the first (40 days) to the third ( 1 20 days) destructive
harvest.
The weight per leaf differed significantly (P<O.O 1 ) during the three destructive harvests
(Table 4-6). In all cases, the control had heavier leaves than the defoliated treatments,
and leaf weight decreased when the intensity of defoliation increased. There were no
significant differences over time.
Chapter Four 89
Table 4-6. Leaf area ratio (LAR), specific leaf area (SLA), weight per leaf (wt.!leaf)
and leaves/gram of Birdsfoot trefoil under different defoliation intensities over
time.
40 days 80 days 120 days SEM I Significance
LAR (mmi/mg) (time x defoliation treatment)
2 cm 2.2 8.7 8A 6 cm 4.6 5.3 7 .7 0.64 1 0 cm 6.2 7.3 6.6 ** Control 5 .8 5 .6 4. 1
SEM (n) OAO (5) 3 .33 (5) OA8 (5) Significance ** NS * *
SLA (mm2/mg)
2 cm 1 2 A 25.7 3 5 .9 6 cm 1 4.2 1 9. 7 29.8 2.54 1 0 cm 1 9.3 28.8 26.2 * Control 1 7.2 20.3 22.2
SEM (n) 1 .39 (5) 3 .24 (5) 3.93 (5) Significance * NS NS
Wt./ leaf (mg)
2 cm 2.3 3 A 2.4 (time) 6 cm 4. 1 3 .8 3 .5 10 cm 5.5 5 . 1 5 . 7 0.48 Control 6.9 7 . 1 8 .2 NS
SEM (n) 0.4 1 (5) OA8 (5) 0.49 (5) Significance * * * * * *
No. Leaves/gram DM
2 cm 267 1 65 1 95 (time) 6 cm 1 58 1 1 1 1 1 1 1 0 cm 1 07 89 85 1 5 . 1 Control 96 77 43 * *
SEM (n) 1 0.8 (5) 9.72 (5) 1 9.3 (5) Si�nificance * * ** * *
* * P < 0.0 1 ; * P< 0.05; NS, not significant; SEM, standard error o f the mean; (n), number of observations for each treatment mean. Note: when time*defoliation treatment interaction was not significant time effect was presented.
Chapter Four 90
The number of leaves per gram differed significantly (P<O.O I ) between treatments,
during the three destructive harvests. In all cases, 2 cm height had a higher number of
leaves than the other treatments (Table 4-6). There was a significant time effect
(P<O.O I ), treatments declining in leaf number per unit of DM from the first to the third
destructive harvest (Table 4-6).
4.4.5 Relationships between herbage harvested and plant components
Herbage harvested after 20 days of growth, following defoliation after destructive
harvests, was correlated with primary roots mass, secondary roots mass, crown size,
root diameter and plant root reserves. Due to reduced association at early stages (first 20
days), results are presented for periods fol lowing defoliation on day 40 and 80. Also
residual leaf area was correlated with herbage harvested, data including the three
periods independently.
4.4.5. 1 The influence of below-ground plant components
In general, association between parameters increased over time (Table 4-7). During the
first period analysed, only primary roots showed a degree of association with herbage
harvested, explaining 29% of growth variation. During the second period (days 80 to
1 00), the primary roots explained 38% of growth variation (Table 4-7).
In contrast, secondary roots mass explained 67% of herbage harvested variation in the
second period, and changes in crown mass explained 66% during final sampling.
However, root diameter did not show any degree of association at any stage (Table 4-7).
Chapter Four 91
Table 4-7. Regressions between BFT herbage harvested (Y, g DM/pot) and below
ground parameters during two periods of growth after defoliation.
Period of growth
Independent variables (x) Day 40 to 60 Day 80 to 100
Primary root mass (glpot) y= 1 .622x+4.267 r2=0.29 P<0.05 Y=1 .584x+5 . 1 5 1 �=0.38 P<0.05
Secondary root mass (glpot) y= 1 . 1 35x+5.44 1 �=0. 1 4 NS y= 0.58 1x+4.577 �=0.67 P<O.O I
Crown mass (glpot) y= I .035x+8.569 �=O.O J NS y= 2.098x+3.732 �=O.66 P<O.O I
Root diameter (mm/root) y= O.65 1 x+4.408 �=O.O2 NS y= O.235x+6.890 �=O.O l NS
4.4.5.2 Residual leaf area for regrowth
A l inear regression was fitted for herbage harvested and residual leaf area during three
periods of growth. In all cases, the increase in herbage harvested was associated with
the increase in residual leaf area (Figure 4-6). Between 40 to 60 and 80 to 1 00 day
periods, variations in residual leaf area explained more than 80% of variation in herbage
harvested, in contrast with initial period (0 to 20 day) that only explained 46% of
variation. Residual leaf area was on average 4 times and 6. 1 times higher for 6 and 1 0
cm treatment respectively than for 2 cm treatment.
Chapter Four
25
20
! � 1 5
1 - 1 0 I 5
o
y (40 to 60)= 0.01 59x + 4.4042
r2 = 0,82
250 500
Residual leaf area (cm2/pot)
750
o
.. 0 to 20 days
o 40 to 60 days
o 80 to l OO days
Linear (0 to 20 days)
-- Linear (40 to 60 days)
-Linear (80 to l OO
y(80 to 1 00) = 0.0 1 56x + 2,7 1 47 ? = 0.86
y(O to 20) = 0,005 1 x + 1 .3693
r2 0.46
1 000
92
Figure 4-6. Regression between residual leaf area and herbage growth of BFT
during three periods (day 0 to 20, 40 to 60 and 80 to 100).
4.4.5.3 Carbohydrate root reserves and regrowth
Total carbohydrate root reserves were significantly correlated with herbage harvested,
re growth increasing when plants had high root reserves (Figure 4-7).
Chapter Four 93
20 y :::= 0.0902x + 1 .0672
� r :::= 0.79 ..... 0 I S � :::s a '-'
1 0 v ..... CIl ...
� 5 0 ... Cl •
0 0 5 0 1 00 1 50 200
Total carbohydrate root reserves (mglplant)
Figure 4-7. Regression between total carbohydrate root reserves and BFT herbage
harvested of pJants defoliated at 20 day intervals (data comprised
average/treatment between day 40 to 60 and 80 to 100).
4.4.5.4 Effect of defoliation intensity on plant components
Information from a multivariate anova (MAN OVA) performed on the main plant
growth components is presented in Table 4-8. From the MANOV A, the first dimension
for defoliation height explained 79, 99 and 99% of variation from destructive harvests
one (40 days) to three ( 1 20 days), respectively. At 40 days, the significance test showed
that defoliation height was not significant when all variables where considered together.
At the second harvest (80 days), the significance test showed differences for defoliation
height (P<0.05), and also root value increased over first harvest, and primary roots and
root diameter were the parameters with highest eigenvectors. At the third harvest ( 1 20
days), there were significant differences for defoliation height (P<O.O l ), increasing the
characteristic root value over previous harvests. Root diameter, secondary roots and
primary shoots were the most affected components (Table 4-8). Thus, the effect of
defoliation increased in significance over time.
Chapter Four 94
Table 4-8. Multivariate analysis of variance (Manova test) performed on
morphological components affected by defoliation height at three successive
destructive harvests (40, 80 and 120 days).
Characteristic Percent Primary Secondary Crown Root Primary Secondary Significance
Root roots roots weight diameter shoots shoots (Wilks' Lambda)
(Eigenvectors)
DESTRUCTIVE HARVEST 1 (40 days)
3 . 1 78.8 -2. 1 4 1 .38 2.89 - 1 .05 5.94 0.53 NS
DESTRUCTIVE HARVEST 2 (80 days)
108.2 99.2 3.54 1 04 1 1 .22 2 .2 1 -0.76 0.5 1
DESTRUCTIVE HARVEST 3 (120 days)
2 1 8.5 99. 1 - 1 .03 3.04 2.58 5.05 -3.34 1 .23 • •
* * P< 0.0 1 ; * P< 0 .05 ; NS, not significant.
4.4.6 Plant survival
Initially, 3 plants were allocated to each pot, and there were no significant differences in
the number of surviving plants per pot between treatments at 40 and 80 days of
evaluation. However, at final destructive harvest ( 1 20 days), the average of plants per
pot was 2.8, 2 .6, 2 .0 and 1 .4 for the undefoliated control, 1 0 cm, 6 cm and 2 cm
defoliation height respectively (SEM 0.303, P<0.05).
4.5 DISCUSSION
The sensitivity of BFT Grasslands Goldie to defoliation intensity demonstrated in field
conditions (Chapter 3) was also explored in this pot trial conducted in parallel. In this
context, plasticity of BFT plants can be understood as "the ability to alter morphology
and physiology in response to varying environmental conditions" (Brandshaw, 1 965;
Chapter Four 95
Schlichting, 1 986; Grime et al., 1 986), to optimise resource capture for re growth after
defoliation. Understanding of these processes required a detailed set of measurements
under controlled glasshouse conditions.
Plant biomass production and the pattern of distribution of components above and
below ground were affected by defoliation intensity. The production of new tissues by
BFT plants after defoliation was higher at lax than hard defoliation within and across
time intervals . During the first cycle of defo liation, RGR increased in BFT plants
defoliated hard (2 cm), but after three cuts differences between intensities disappeared
and RGR remained unchanged over time (Figure 4-2). These results showed that initial
plant adjustments to compensate the amount of biomass removed increased the relative
regrowth rate (Oesterheld, 1 992; Richards, 1 993). However, there are limitations to the
availability and allocation of resources to maintain this adjustment.
Intensively defoliated plants tended to allocate relatively more metabolites to the
production of new leaf area, as observed in LAR and SLA relationships and in the
number of leaves per unit of biomass (Table 4-6). Plants defo liated intensively tended to
have a dense canopy of very small leaves (Table 4-6). This is a common phenomenon in
temperate legumes, leaf size being highly sensitive to changes in defoliation patterns
(Chapman and Lemaire, 1 993). Treatment differences in residual leaf area explained
differences in regrowth, the greater the residual leaf area the higher the re growth
(Figure 4-6).
Root carbohydrate reserves contributed to differences in growth rate. Starch declined
fol lowing initial defoliation (Table 4-3), suggesting that these compounds were
mobilised and used for successive growth, but reserves were not restored in the plants
defoliated regularly. In contrast, control plants growing under undisturbed conditions
stored starch and free sugars, particularly after 80 days growth accompanied by a
substantial root development. Very low carbohydrate reserves in BFT during the active
growth season were reported by Smith ( 1 962), reserves not being fully replenished after
successive defoliation. This is a clear difference from lucerne or red clover, where
levels of accumulation and patterns of utilization of root reserves are clearer (Smith,
1 962). In this trial, the frequent defoliation imposed (each 20 days) reduced the
Chapter Four 96
opportunity for replenishing reserves, as occurs when defoliation interval is extended
from 2 1 to 42 days (Alison and Hoveland, 1 989 b). Thus, re growth was more dependant
on residual l eaf area than root reserves. Root reserves are considered important in
legume survival, particularly in those cases when plants are under stress (lung and
Smith, 1 96 1 ). The starch content and its conversion to soluble sugars are believed to
contribute to increased winter hardiness of crown forming species (Bula et al., 1 956),
despite the existence of other root constituents that can contribute to winter hardiness
(Boyce and Volenec, 1 992; Li et al., 1 996). Accumulation of carbohydrates in BFT is
regulated by photoperiod and temperature, reserve accumulation patterns increasing
under short days and low temperatures in autumn (Nelson and Smith, 1 969; Greub and
Wedin, 1 97 1 ). Seasonal accumulation patterns of root reserves explain why BFT
tolerates frequent but not close defoliation. Carbohydrates for regrowth depend
primarily upon residual leaf area because root reserves are low (Nelson, 1 995).
Production of above-ground biomass over 1 20 days of the undefoliated control was
higher than any of the defoliated treatments, and the hard defoliated treatment the least
productive. The cumulative effect of defoliation intensity over time was observed on
above and below-ground plant components (Table 4-7), effects being significant from
80 days onwards. Below-ground biomass was reduced under repeated and severe
defoliation (2 cm), moderately affected under 6 cm height, and was little affected or
even increased on the 1 0 cm treatment, compared with plants growing undisturbed (see
also Pierre and Jackobs, 1 95 3). Crown size was negatively associated with the intensity
of defoliation. It is suggested that reduction in crown size affects plant reserves, as
reported by Greub and Wedin ( 1 97 1 ). At the end of the experiment, the below-ground
parameters most affected by defoliation treatment were root diameter and secondary
root mass (Table 4-8). It is suggested that excessive defoliation during spring could
severely affect plant survival, partiCUlarly in the fol lowing summer. An efficient root
system is defined by the degree of exploration of soil (Barber, 1 974 cited by Alison and
Hoveland, 1 989 b). These results showed an intense decline in secondary roots mass
with increasing defoliation severity, suggesting that BFT plants severely defoliated
reduced capacity to capture water and nutrients, and finally reduced the degree of plant
survival .
Chapter Four 97
Root diameter of BFT was affected by defoliation intensity. Alison and Hoveland ( 1 989
b) observed similar effects of defoliation frequency in some BFT cultivars, but effects
were not consistent. In this experiment, herbage growth was not correlated with root
diameter of BFT Grasslands Goldie after 40 or 80 days. After 1 20 days, root diameter
was the parameter that showed the closest association to defoliation intensities applied.
However, root mass was a better predictor of herbage growth.
Overall, the intensive and repeated defoliation of BFT plants produced a general decline
in below-ground mass that influenced re growth after defoliation. Plants made some
short term morphological adjustments but these were ineffective under hard defoliation
partially attributed to the reduced potential for storage of root reserves during active
growth season in comparison with other crown-forming species l ike alfalfa or red clover
(Smith, 1 962; Smith and Nelson, 1 967). In fact, BFT shows a l imited morphological
and physiological plasticity to defoliation, as indicated by the increase in RGR of
severely defoliated plants observed during the first cycle of regrowth. The absence of
morphological responses does not mean plants lack plasticity, because stability can be
achieved by altering only simple factors like growth rate (Schlichting, 1 986). A high
degree of morphological plasticity could be defined as an advantage if large amounts of
resources are available, and also the degree of adaptation to stress conditions shows
reduced changes in morphology and conservative utilization of resources (Grime et al.,
1 986).
BFT plants were more dependent on residual leaf mass, and moderate to lax defoliation
intensities showed advantages in herbage production and plant surviva1 . Comparing 6 to
1 0 cm stubble height, 1 0 cm showed the best production and plant survival, though
other factors like the frequency of defoliation should be considered. There was an
increase in density of primary shoots on plants defoliated at 6 cm in comparison with
those defoliated at 2 or 1 0 cm after 1 20 days (Table 4-5). This tendency was also
observed in results presented in Chapter 3, where plants defoliated at 6 cm showed
more primary shoots per plant at the end of the spring (see Table 3-4 to draw individual
plant comparisons). The secondary shoots were increased at early stages (40 and 80
days) by 6 cm defoliation (Table 4-5), but at late stages studied ( 1 20 days) plants
defoliated laxly ( 1 0 cm) showed more shoots than those defoliated more intensively. As
Chapter Four 98
reported for lucerne, shoots development is associated with the intensity of defoliation
(Keoghan, 1 970) and probably, based on these results, with the extension of the period
under defoliation. The lax defoliation could determine a higher importance of secondary
shoots by a high stubble height, reducing the potential of development of new shoots.
The hard defoliation could stimulate the formation of new shoots, but plant reserves will
be limiting shoots development if intensive defoliation is repeated over time. Moderate
defoliation intensities could suggest a positive effect of defoliation in the formation of
new shoots, as well as the importance of secondary shoots.
4.6 CONCLUSIONS
Vigour of BFT Grasslands Goldie plants was markedly reduced by regular close
defoliation. Root reserves declined during the active growing season, re growth being
more dependent on residual leaf area. Root mass and carbohydrate reserves were also
diminished. These factors l imited regrowth potential, and eventually reduced plant
survival. The evidence showed that Grasslands Goldie is sensitive to close defoliation,
despite the reported semiprostrate plant habit. BFT plasticity in response to intensive
defoliation comprised increase in RGR, LAR, SLA and production of new leaves to
compensate the loss of tissues. However, these adjustments were not enough to
compensate the amount of tissues removed by cutting. A stubble height of 1 0 cm after
defoliation was the best combination for plant production and survival when plants were
frequently defoliated. These results obtained in controlled glasshouse environment,
were consistent with those observed in field conditions (Chapter 3), but provided a more
detailed explanation of morphological and physiological processes of defoliated plants.
Chapter Four
4.7 REFERENCES
99
ALISON, M.W. AND HOVELAND, C.S. ( 1 989 a). B irdsfoot trefoil management. H. Yield, quality and stand evaluation. Agronomy Journal S1 : 745-749.
ALISON, M.W. AND HOVELAND, C.S. ( 1 989 b). Birdsfoot trefoil management. 1. Root growth and carbohydrate storage. Agronomy Journal S1 : 739-745 .
BOLOGNA, 1.1. ( 1 996). Studies on strategies for perennial legume persistence III
lowland pastures. Thesis of Master of Agricultural Science. Lincoln University,
New Zealand. 220 p
BOYCE, P.1. AND VOLENEC, 1.1. ( 1 992). Taproot carbohydrate concentrations and
stress tolerance of contrasting alfalfa genotypes. Crop Science 32: 757-76 1 .
BRANDSHA W, A.D. ( 1 965). Evolutionary significance of phenotypic plasticity in
plants. Advances in Genetics 13 : 1 1 5- 1 55 .
BULA, R.1., SMITH, D. AND HODGSON, H.J. ( 1 956). Cold resistance and chemical
composition in overwintering alfalfa, red clover and sweetclover. Agronomy
Journal 46: 397-40 1 .
CHAPMAN, D.F. AND LEMAIRE, G. ( 1 993). Morphogenetic and structural
determinants of plant regrowth after defoliation. Proceedings of the XVlI International Grassland Congress. New Zealand and Australia. pp. 95- 1 04.
CORDEIRO DE ARAUJO, 1. AND JACQUES, A.V.A. ( 1 974). Influencia do estadio
de crescimento e da altura de corte sobre as reservas de glicidos e nitrogenio
total de cornichao (Lotus corniculatus L.). (Effect of physiological stage and
height of cutting on sugar reserves and nitrogen in birdsfoot trefoil). Revista da
Sociedade Brasileira de Zootecnia. Vol. 3(2): 1 23 - 1 37 .
FANKHAUSER Jr., J.1. AND VOLENEC, J.1. ( 1 989). Root vs . shoot effects on
herbage re growth and carbohydrate metabolism of alfalfa. Crop Science 29:735-
740.
FEL TNER, K. AND MASSENGALE, M.A. ( 1 965). Influence of temperature and
harvest management on growth, level of carbohydrates in roots, and survival of
alfalfa (Medicago sativa L.). Crop Science 5: 585-588.
Chapter Four 100
FRAME, J . ; CHARLTON, IF.L. AND LAID LAW, A.S. ( 1 998). B irdsfoot trefoil and
greater lotus. In Temperate Forage Legumes. CAB International, Wallingford.
ISBN 0-85- 1 99-2 1 4-5. Chapter 6. pp. 245-27 1 .
GRAB ER, L.F. ; NELSON, N.T. ; LUEKEL, W.A. AND ALBERT, W.B. ( 1 927).
Organic food reserves in relation to the growth of alfalfa and other perennial
herbaceous plants. University of Wisconsin Agriculture Experimental Station.
Bulletin 80.
GREUB, LJ. AND WEDIN, W.F. ( 1 97 1 ). Leaf area, dry matter accumulation and
carbohydrate reserve levels of birds foot trefoil as influenced by cutting height.
Crop Science 1 1 : 734-738.
GRIME, J.P., CRICK, lC. AND RINCON, lE. ( 1 986). The ecological significance of
plasticity. Plasticity in Plants. Edited by Jennings, D.H. and Trewavas, AJ. The
company of Biologists Lmtd., Cambridge. pp.5-30.
HODGKINSON, K.C. ( 1 968). Studies of the physiology of regeneration of lucerne
(Medicago sativa L.). Review of Ph.D. Thesis University of New England. New
South Wales 1 967. Journal of Australian Agricultural Science 34: 22 1 .
HODGSON, J. ( 1 979). Nomenclature and definitions in grazing studies. Grass and
Forage Science 34: 1 1 - 1 8 .
JUNG, G.A. AND SMITH, D. ( 1 96 1 ). Trends of cold resistance and chemical changes
over winter in the roots and crowns of alfalfa and red clover. n. Changes in
mineral constituents. Agronomy Journal 53 : 363-366.
LI, G. ( 1 997). Response of Chicory (Cichorium intybus L.) to defoliation. Thesis
presented for the degree of Doctor of Philosophy in Plant Science. Massey
University, New Zealand. 1 97 p
NELSON, C .J. ( 1 995). Photosynthesis and carbon metabolism. In Forages. Vol. 1 . An
Introduction to Grassland Agriculture. Edited by R.F. Barnes, D.A. Miller and
C.J. Nelson. Fifth edition. pp. 3 1 -43 .
NELSON, CJ. AND SMITH, D. ( 1 968). Growth of birdsfoot trefoil and alfalfa. Ill.
Changes in carbohydrate reserves and growth analysis under field conditions.
Crop Science 8: 25-28.
NELSON, C.J . AND SMITH, D. ( 1 969). Growth of birdsfoot trefoil and alfalfa. IV.
Carbohydrate reserve levels and growth analysis under two temperature regimes.
Crop Science 9: 589-59 1 .
Chapter Four 101
OESTERHELD, M. ( 1 992). Effect of defoliation intensity on above-ground and below
ground relative growth rates. Oecologia 92: 3 1 3-3 1 6.
PIERRE, II AND JACKOBS, lA. ( 1 953) . The effect of cutting treatments on
birdsfoot trefoil . Agronomy Journal 45: 463-468.
RICHARDS, 1.H. ( 1 993). The physiology of plants recovering from defoliation.
Proceedings of the XVII International Grassland Congress. New Zealand and
Australia. pp. 85-94.
SIL V A, 1.P. ( 1 968). Interrelations of leaf area and carbohydrate root reserves as
determinants of the regrowth potential of alfalfa (Medicago sativa L.)
Dissertation Abstracts 29, 1 906.
SCHLICHTING, C.D. ( 1 986). The evolution ofphenotypic plasticity. Annual Review of
Ecological Systems 17 : 667-693 .
SMITH, D. ( 1 962). Carbohydrate root reserves in alfalfa, red clover, and birdsfoot
trefoil under several management schedules. Crop Science 2 : 75-78.
SMITH, D. AND NELSON, C.J. ( 1 967). Growth of birdsfoot trefoil and alfalfa. I .
Responses to height and frequency of cutting. Crop Science 7: 1 30- 1 33 .
Chapter Five 102
5. PERFORMANCE OF FOUR Lotus corniculatus L.
CULTIVARS IN RESPONSE TO INTENSITY AND
TIMING OF DEFOLIATION
5 . 1 ABSTRACT
5.2 INTRODUCTION
5 .3 MATERIALS AND METHODS
5 .3 . 1 Measurements
5 .3 .2 Statistical analysis
5 .4 RESULTS
5 .4. 1 Climate conditions during experimental period
5 .4.2 Herbage production to first harvest
5 .4.3 Annual herbage production
5 .4.4 Growth rates
5 .4.5 Sward characteristics
5 .4.6 Nutritive value
5 .4.7 Plant density
5 .4.8 Plant morphology
5 .5 DISCUSSION
5 .6 CONCLUSIONS
5.7 REFERENCES
Chapter Five
5.1 ABSTRACT
103
A field experiment was conducted from May 1 998 to April 2000 at INIA Treinta y Tres,
Uruguay, South America to study the effects of intensity of defoliation and timing of
initial defoliation on herbage production and morphological adaptations of four
cultivars of Lotus corniculatus L. (birds foot trefoil, BFT). A factorial experiment
(4x2x3) was applied in a complete randomised block design with four replicates. Tested
factors included four BFT cultivars (San Gabriel, INIA Draco, Grasslands Goldie and
Steadfast), two defoliation intensities (4, 8 cm height) and three times of initial
defoliation during the establishment year (vegetative, 50% flowering and advanced
maturity). During the second year, the original defoliation intensities treatments were
each split further into two intensities (4 and 8 cm height). Plants were defoliated from
mid spring to early autumn at intervals of 40 days each year, but cultivars were
managed in two groups defined as local (San Gabriel and INIA Draco - winter active)
and introduced (Grasslands Goldie and Steadfast - winter dormant). Herbage
production, sward height, nutritive value, plant density and plant morphology were
recorded. Local cultivars were 2.6 and 2.5 times more productive than introduced
cultivars during Year 1 (6.3 vs. 2.4 t DM/ha) and Year 2 (4.7 vs. 1 .8 t DM/ha)
respectively, with production advantage throughout the year. All cultivars showed
adequate standards of forage quality, varying from 590-720 g/kg DM for digestibility of
organic matter, 25-39 g/kg DM for nitrogen, 230-400 g/kg DM for acid detergent fibre
and 22-3 1 g/kg DM for condensed tannins. Plant density decreased drastically during
the second year as a consequence of severe drought conditions. There were substantial
differences in morphology between cultivars, especially in root and crown size and
shoots per plant. Production was greater when defoliated at 4 than 8 cm height for all
cultivars. The timing of initial defoliation affected annual production on Year I , but
only minor effects were observed in Year 2.
Keywords: Lotus corniculatus L. cultivars, defoliation management, forage production,
nutritive value, plant morphology, plant population.
Chapter Five
5.2 INTRODUCTION
104
Birdsfoot trefoil (BFT) is a widely used forage legume in extensive and intensive
farming systems of Uruguay, growing alone or in mixtures with grasses and other
legumes. Its success is recognised in a wide range of environments and soil conditions,
for forage and seed production. Lack of persistence is reported, due to inadequate
defoliation strategies, and high incidence of root and crown diseases with production
and population declining from the third year of use (Altier, 1 988; Rebuffo and Altier,
1 996).
Despite these limitations, BFT is an important summer feed alternative in pastoral areas
where especially feed quality and sometimes feed quantity are l imited. Thus, potential
growth, levels of accumulation and associated nutritive value of BFT are key issues in
production systems.
The objectives of this research were to explore the degree of adaptability, productivity
and nutritive value of contrasting BFT cultivars in the eastern region of Uruguay, and to
investigate more effective defoliation strategies for the species. The effects of
management decisions like timing and intensity of defoliation in the year of
establishment on herbage production, and on the morphology and physiology of BFT
plants, were examined.
Four BFT cultivars with contrasting characteristics were selected to explore defoliation
responses, two commercial cultivars used in Uruguay (San Gabriel and INIA Draco),
one from New Zealand (Grasslands Goldie) and one from the United States (Steadfast).
Introduced from Brazil around 50 years ago (Gardner et al. , 1 968), BFT San Gabriel is
the oldest cultivar used in Uruguay. It shows a high degree of adaptability to the local
ecological conditions. BFT Ganador, an Uruguayan selection, is another recommended
cultivar (Canimbula et al., 1 996), but productivity of these two cultivars has reduced
Chapter Five 105
drastically by diseases in areas where they have been used extensively (Altier, 1 997).
Formoso ( 1 993) reported values of 4, 1 0, 7 and 5 t DMlha from the first to fourth year
of production, working with BFT San Gabriel. Digestibility and nitrogen values are
high in early spring, 730 and 35 g/kg of DM respectively, but decline over the growing
season to 6 1 0 and 30 g/kg ofDM respectively in early autumn (Formoso, 1 993).
Research efforts in Uruguay are oriented to increase productive persistence by selecting
new cultivars resistant to disease complexes that can substitute traditional cultivars.
Recently, the cultivar INIA Draco was developed, fol lowing two cycles of selection of
field persistent plants from parental material of BFT Ganador and a local population
from La Estanzuela, Uruguay, with the objective to extend pasture l ife to 3-4 years and
expand the range of cultivars available (Rebuffo and Altier, 1 996). Advantages in the
forage production of BFT INIA Draco of 8, 12 , 42 and 74% over BFT San Gabriel were
observed from the first to fourth year of pasture respectively (Rebuffo and Altier, 1 996).
Crown size, density of stems and leaf proportion are improved characteristics of INIA
Draco. In general, Uruguayan cultivars have intermediate to erect growth habits, are
early flowering and are winter active (F ormoso, 1 993 ; Rebuffo and Altier, 1 996). The
nutritive value of BFT cultivars need to be more studied, in particular for those recently
developed. Also, there are no references about tannin content in the conditions of
Uruguay, levels of concentration of these compounds having significance in animal
performance.
BFT Grasslands Goldie is described as a semi-prostrate cultivar adapted to grazing
(AgResearch, 1 995), with winter dormancy in the conditions of Uruguay (Juan
Bologna, personal communication). Steadfast is a prostrate cultivar, with small to
medium sized leaves and stems (Norberg, 1 999), and is the first BFT cultivar with the
ability to spread by rhizomes. It was developed from the mating of accessions from
Morocco with commercial cultivars Norcen and AU-Dewey and germplasm MU-8 1
from USA (Beuselinck and Steiner, 1 995). Early reports showed forage production of
3 . 5 t DM/halyear in Iowa (Norber, 1 999), which was lower than the traditional BFT
cultivars planted that produced between 4.4-4.9 t DMlha. There are no reports on the
performance of Steadfast in Uruguay.
Chapter Five
5.3 MATERIALS AND METHODS
1 06
The study was sited at Palo a Pique Research Unit, INIA, Treinta y Tres, Uruguay
(latitude 3 3° 54' S , longitude 54° 3 8 ' W), on a fine, thermic, mixed, vertic Argiudoll
(ARS-USDA classification, Femando Garcia, personal communication) moderately
fertile, poor in phosphorus content and acid (Table 5 - 1 ) .
Table 5-1 . Soil characteristics at experimental site in Palo a Pique Research Unit.
(Source: Laboratory of Soils, INIA La Estanzuela, Uruguay).
Soil testing dates Soil depth pH Organic carbon P (Bray I ) K (cm) (H2O) (%) (�g Pig soiQ (mequiv/l OOg soil)
April 1 998 0 - 7.5 5 .4 2.6 3.0 0.3 7.5 - 1 5 5 .5 1 .6 2 .9 0 .2
March 1 999 0 - 7.5 5.5 2.5 6.5 0.3 7.5 - 1 5 5 .6 1 .4 2.3 0.2
The trial was sown manually over a conventional seedbed on May 8 1 998, and
evaluated to April 2000. The BFT cultivars established were San Gabriel (Brazil),
INIA Draco (Uruguay), Grasslands Goldie (New Zealand) and Steadfast (USA). The
seeding rate of 1 2 kglha was based on that of INIA Draco and corrected for germination
and purity and adjusted for seed weight to sow the same number of viable seeds per
cultivar (870 viable seeds/m2). Seed was inoculated with Rhizobium and pelleted.
Superphosphate (N-Ptotal -Psoluble-K; 0-2 1 -23-0) was applied at 26 kg Plha at seeding
time, fol lowed by 26 kg Plha in March 1 999.
A factorial experiment (2x3x4) was used with treatments laid out in plots of 5 x 2.5 m
in a complete randomised block design with four replicates (Plate 5 - 1 ). Treatments were
a combination of two intensities of defoliation (4 or 8 cm cutting height) and three
initial defoliation times in the year of establishment (defined as vegetative, 50%
flowering and advanced maturity), applied to the four BFT cultivars previously
described.
Chapter Five 107
Because of differences in the growth pattern of the cultivars, they were managed in two
independent groups. The first group comprised the cultivars with winter activity (San
Gabriel and INIA Draco, defined as local cultivars). Their cutting sequence started on
November 4 (vegetative), December 1 5 (50% flowering) and January 25 (advanced
maturity). The second group was the winter dormant cultivars (Grasslands Goldie and
Steadfast, defined as introduced cultivars), for which defoliation commenced one
month later on December 4 (vegetative), January 1 3 (50% flowering) and February 22
(advanced maturity). After the initial cut, defoliation was at intervals of 40 days until it
ceased in April 1 999 (April 5 and 1 5 for introduced and local cultivars, respectively),
during the first year (Table 5-2). After that, the trial was uncut until October 4, 1 999 for
local cultivars, and November 1 5, 1 999 for introduced cultivars. At this time, original
plots were split into sub-plots 2 .5 m x 2.5 m, and two defoliation intensities (4 cm and 8
cm) were applied over the original defoliation intensities in a standard defoliation
routine each 40 days to April 2000. The initial defoliation time treatments were not
repeated during the second cutting season (Table 5 -2).
Table 5-2. Cutting schedule of BFT plots sown on May 1998.
Year Cultivars Initial time (year I ) Dates of cutting
4 Nov 1 5 Dec 25 Jan 8 Mar 1 5 Apr
Vegetative x x x x x
Local 50% Flowering x x x x
Advanced maturity x x x
Year I 4 Dec 1 3 Jan 22 Feb 5 Apr
( 1 998-99) Vegetative x x x x
Introduced 50% F lowering x x x
Advanced maturity x x
4 0ct 1 5 Nov 24 Dec 1 Feb 1 0 Mar 20 Apr
Local x x x x x x
Year 2
( 1 999-00)
Introduced x x x x x
Chapter Five 108
When specifically referred, the seasons over the year were defined as autumn (from
March to May), winter (from June to August), spring (from September to November)
and summer (December to February).
5.3.1 Measurements
5.3.1 .1 Forage production
Forage production was measured by cutting to defined sward heights a 1 m x 5 m strip
for initial plots and 1 m x 2.5 m for split plots, using a reciprocating blade machine. Ten
sward heights were recorded in each plot before cutting, using a ruler. Forage samples
were weighed fresh in the field and subsamples taken for dry matter and botanical
composition. One subsample was oven-dried at 60 QC for 48 hours for dry matter and
forage quality determinations, and another subsample was separated into components
(Birdsfoot trefoil, grasses and weeds). Samples of the botanical fractions were oven
dried to determine dry matter content.
Prior to cutting, a 1 00 mm x 100 mm quadrat was cut to ground level, and separated
into strata mass above and below cutting height treatment (4 or 8 cm) to estimate total
leaf area and leaf fractions above and below cutting height. Leaves were dissected from
stems, and area measurements made using a LI-Cor LI-3 1 00 Leaf Area Meter (Lambda
Instruments Co, Lincoln, NE, USA). Leaf/stem ratio of each sample was estimated after
oven-drying dissected leaves and stems.
5.3. 1 .2 Forage quality
Determinations of in vitro organic matter digestibility (OMD) (Til ley and Terry, 1 963),
nitrogen by Micro-Kjeldahl and acid detergent fibre (ADF) (Goering and Van Soest,
1 970) were made for the three initial defoliation times in the two groups of cultivars.
After first time defoliation, quality parameters were monitored regularly at 40 day
intervals during the cutting season. The accumulation of digestible organic matter
Chapter Five 109
(DOMA= herbage dry matter accumulation, corrected for ash content, multiplied by
organic matter digestibility) was calculated for the different physiological and presented
for San Gabriel and INIA Draco. Determinations were made at the Laboratory of
Nutrition of INIA La Estanzuela, Uruguay.
5.3. 1 .3 Condensed tannins
Two forage samples of each cultivar were hand clipped at the vegetative stage
(September, 1 999), stored at -20 QC, dried and ground to pass through a 1 mm sieve.
Condensed tannins (CT) fractions (acetone/water-extractable, protein-bound and fibre
bound) were determined using the butanol-HCL method of Terril l et al. ( 1 992).
Analyses for extractable and protein-bound fractions were made on two sub-samples,
and for fibre-bound fraction in one sample per treatment. CT determinations were
conducted at the Institute of Food Nutrition and Human Health, Massey University,
Palmerston North (Juan Mieres, personal communication).
5.3.1 .4 Plant density and morphology
Plant density of BFT was counted in July 1 998, March 1 999, October 1 999 and March
2000 by collecting a soil block 250x250x250 mm in each plot. Plants were washed to
remove soil, and counted. Then, 5 BFT plants were randomly taken for morphological
measurements. In each plant, the number of primary shoots, secondary shoots,
rhizomes, new developing shoots, primary root length and diameter were recorded.
Crown and roots (> 2 mm diameter) were oven-dried at 60 QC for 48 hours for dry
weight.
Chapter Five 1 1 0
5.3.1 .5 Plant architecture
In December 1 999, the vertical distribution of plant tissues within the BFT sward
canopy after 30 days of re growth was analysed, using an inclined point quadrat
(Warren Wilson, 1 963), placed at 32.5° to the horizontal. Ten stations per treatment
were recorded, and in each case contacts included living structures (leaves and stems),
dead structures (leaves and stems) and reproductive structures (flowers and pods) of
BFT and other species. Information was expressed by the number of contacts per 4 cm
of sward height, and set out graphically to determine the distribution of different
components within the sward.
5.3.2 Statistical analysis
Data were analysed using the statistical package SAS (SAS, 1 990). The analysis of
annual herbage production was based on a factorial design with 4 replicates for Year 1 ,
with cultivars, defoliation intensity and time of initial defoliation as the main factors. In
Year 2, the factor defoliation intensity in Year 1 was split in two defoliation intensities
for Year 2, with defoliation height of Year 1 as the main plot, and defoliation height of
Year 2 as the split-plot factor.
The differences between cultivars determined that they were managed m two
independent pairs. Thus, accumulation dates, growth rates and forage quality
parameters were recorded at different dates for each pair. The statistical analysis of
these parameters was made for each group independently (local or introduced
cultivars), using a 'repeated measures ' option of SAS GLM procedures. Total
condensed tannins, protein-bound and extractable fractions were analysed using a
complete randomised design with two replicates, but fibre-bound fraction was not
statistically analysed because analyses were not replicated.
Variation in plant density over time was examined by 'repeated measures' analysis, and
morphology parameters at the end of each year as independent dates, in both cases for
all cultivars simultaneously. Also, morphological parameters were all analysed
Chapter Five 1 1 1
simultaneously by canonical discriminant analysis using the CANDISC procedure of
SAS programme (SAS, 1 990).
Plate 5-1 . General view of BFT cultivars under different intensities and timing of
defoliation in the year of establishment (Year 1).
Chapter Five
5.4 RESULTS
5.4. 1 Climate conditions during the experimental period
1 12
Rainfall was 24% higher during 1 998 and 29% less during 1 999 than the 8-year
average. During winter rainfall was 30% higher than average, but during spring and
summer it was lower, especially in 1 999 with rainfall 52 and 69% less than average
respectively (Table 5-3). From the end of spring 1 999 to summer 2000, soil
temperatures were higher than average (Table 5-3).
Table 5-3. Monthly rainfall and soil temperature at 50 mm depth during the
evaluation period and the 8-year average (Source: INIA, Treinta y Tres, Uruguay).
Rainfall (mm) Soil Temperature eq ] 998 ] 999 2000 8-year 1 998 1 999 2000 8-year
average average
January ] 47 48 66 97 24.6 27.2 30 . 1 27.9 February 58 1 4 1 28 1 03 25 .7 27 . 1 27.6 26.5 March 97 1 39 1 20 1 04 22.6 26.9 25 . 1 24.9 April 288 87 1 3 7 20.2 1 8.7 1 9.9 May ] 28 46 1 00 1 5 .8 1 5 .0 1 5 .9 June 1 68 259 1 23 1 2.6 1 2 . 1 1 2.5 July 2 1 3 6 1 1 09 1 3 .5 1 2.3 1 2.0 August 99 68 1 0 1 1 4.0 1 3. 8 1 3.9 September 57 64 84 1 6.5 1 6.9 1 6.3 October 63 40 90 2 1 .5 1 9.8 20.2 November 7 1 1 9 8 1 23.9 24.4 24.3 December 1 33 26 1 04 25 .5 27.7 27.2
Total 1 523 870 -- 1 232 -- -- -- --
Soil water balance was positive during autumn and winter 1 998 and part of winter 1 999
(June and July). During the rest of the evaluation period water balance was negative,
severely so from October 1 999 to February 2000 (Figure 5 - 1 ). From November 1 999 to
January 2000, the water deficit was between 1 3 0 and 1 47 mm.
Chapter Five
250
200
1 50
100
� 50 l!l � 0 :.:::: co
� -50 r! . <C -100 I
I -150 � -200 l -250 l
c::::J Effecti'>El Rainfall _ EVP - Soil water balance
o 9 c (Il ...,
1 13
Figure 5-1 . Effective rainfall, evapotranspiration (EVP) and soil water balance for
the period April 1 998-February 2000 in Palo a Pique, Research Unit (Raid
Bermudez and Jose Terra, personal communication).
5.4.2 Herbage production to first harvest
Three initial defoliation times were defined (vegetative, 50% flowering and advanced
maturity), that differed between the two pairs of cultivars, because of the different
periods of forage accumulation from sowing to first harvest. Local cultivars had a faster
initial growth than introduced materials, so were defoliated earlier. Analysis was
therefore done for BFT fraction, because other components made a minimum
contribution, and for each group independently.
Chapter Five 1 14
Table 5-4. Herbage harvested (kg DM/ha) and plant height at harvest (cm) of local
BFT cultivars from sowing (8/5/98) to three different stages of development in the
establishment year.
Herbage harvested (kg DMlba)
Cultivars
Defoliation height
INIA Draco San Gabriel SEM (n) Significance
4 cm 8 cm SEM (n) Significance
Plant height at harvest (cm)
INIA Draco San Gabriel SEM (n) Significance
Vegetative 4 November ( 1 80 days)
700 1 630
1 05 (8) **
1 380 950
1 05 (8) *
20 29
1 .3 (8) **
50% F lowering 1 5 December (22 1 days)
2370 2690 56 (8)
**
2850 22 1 0 5 6 (8)
* *
3 3 3 8
0 . 4 (8) **
Advanced maturity 25 January (262 days)
3360 3360
1 32 (8) NS
3630 3090
1 32 (8) *
3 7 43
0.6 (8) * *
* *, P<O.O 1 ; *, P<O.05; NS, not significant; SEM, standard error o f the mean; (n), number of observations for each treatment mean
For local cultivars, there was a significant time x cultivar effect (P<O.O l ), accumulation
increasing over time. There were significant differences between cultivars and
defoliation height (Table 5-4), but no significant interaction between these variables.
San Gabriel accumulated significantly (P<O.O 1 ) more dry matter than INIA Draco up to
November and December, but by January herbage harvested for the two cultivars was
identical . The height of defoliation determined levels of dry matter harvested, in all
cases 4 cm height producing more than 8 cm height. San Gabriel showed significantly
(P<O.O l ) greater plant height than INIA Draco at all dates. (Table 5 -4). Maximum plant
heights at first harvest in January (262 days after sowing) were 43 and 37 cm for San
Gabriel and INIA Draco, respectively.
Chapter Five 1 1 5
Table 5-5. Herbage harvested (kg DM/ha) and plant height at harvest (cm) of
introduced BFT cultivars from sowing (8/5/98) to three different stages of
development in the establishment year.
Herbage harvested (kg DM/ha)
Cultivars
Defoliation height
Grasslands Goldie Steadfast SEM (n) Significance
4 cm 8 cm SEM (n) Significance
Plant height at harvest (cm)
Grasslands Goldie Steadfast SEM (n) Significance
Vegetative 4 December (2 1 0 days)
760 1 060
75 (8) *
1 290 530
75 (8) * *
22 20
0.6 (8) NS
50% Flowering 1 3 January (250 days)
1 720 1 580
1 19 (8) NS
2 1 30 1 1 80
1 1 9 (8) * *
3 0 25
0.5 (8) * *
Advanced maturity 22 February (290 days)
1 850 1 950
1 20 (8) NS
2660 1 1 30
1 20 (8) * *
3 2 26
0.9 (8) * *
** , P<O.O I ; * , P<0.05 ; N S , not significant, SEM, standard error of the mean; (n), number of observations for each treatment mean
For introduced cultivars, there was a significant interaction time x height of defoliation
effect (P<O.O 1 ) . Only in early December, there was a significant difference (P<0.05) in
herbage harvested in favour of Steadfast (Table 5-5). The height of defoliation affected
herbage harvested (P<O.O 1 ), in all cases the yield was higher when defoliated at 4 cm
rather 8 cm. Plant height showed significant differences in January and February
(P<O.O l in both cases), Grasslands Goldie being taller than Steadfast (Table 5-5).
Maximum plant heights at final harvest (290 days from sowing) were 32 and 26 cm for
Grasslands Goldie and Steadfast respectively.
Chapter Five 1 1 6
5.4.3 Annual herbage production
The annual herbage production of Year 1 ( 1 998- 1 999) and Year 2 ( 1 999-2000) was
analysed for each pair of cultivars (local and introduced), because of differences in
growth patterns. Comparisons between the two pairs are also presented.
Year 1
The adequate plant establishment limited the incidence of other species, mainly some
annual grasses ( Vulpia australis and Gaudinina fragilis). The contribution of these
species for local cultivars was 0. 1 3 t DM/ha, and for introduced cultivars 0.25 t DM/ha.
There were no differences between cultivars in each group, so only results for the BFT
contribution are presented. The overall herbage production of BFT from sowing in May
1 998 to April 1 999 (Year 1 ) was 4.3 t DM/ha. The average production of the winter
active BFT (local cultivars) was 2.6 times higher than the average of winter latency
BFT (introduced cultivars) (6.3 vs 2.4 t DM/ha for local and introduced cultivars
respectively, SEM 0. 1 6, P<O.O 1 ). There was a significant interaction (P<0.05 in both
groups) cultivar x initial defoliation time (Figure 5-2).
8
o
Local cultivars
]
4-Nov I S-Dec 2 S -Jan
Introduced cultivars
I
4-Dec l 3-Jan 22-Feb
T ime of first defoliat io n
Gabriel � INIA [)race . Grasslands Geldie 11 �te�dfa��
Figure 5-2. Annual herbage production (t DM/ha/year) of BFT cultivars affected
by timing of initial defoliation during the Year 1. Vertical bars represent SEM for
each group of cultivars, (n=8).
Chapter Five 1 1 7
Between local cultivars, the highest production was achieved by San Gabriel when
cutting started in November or December, and by INIA Draco when cutting started in
December (Figure 5 -2). For introduced cultivars, Steadfast showed the highest
production when the initial cut was done in December and Grasslands Goldie when
cutting sequence started on January (Figure 5 -2).
For local cultivars, production was increased 23% when defoliated at 4 cm rather than 8
cm height (6.9 and 5 .6 t DM/halyear for 4 and 8 cm height respectively, SEM 0. 1 5 ,
P<O.O I ). The results for introduced cultivars showed a 1 00% increase when defoliated
at 4 rather 8 cm height (3.2 and 1 .6 t DM/halyear for 4 and 8 cm height respectively,
SEM 0. 1 0, P<O.O I ).
Year 2
The average of herbage production from April 1 999 to April 2000 (Year 2) was 3 .3 t
DM/halyear. As reported in Year 1 contribution of other species was low, 0. 1 and 0.36 t
DM/ha in average for local and introduced cultivars respectively, information is
presented for BFT fraction only. The production of local cultivars was 2.5 times
(P<O.O 1 ) greater than introduced cultivars (Table 5 -6).
Between local cultivars, San Gabriel and INIA Draco did not differ in production, the
average being 4 .7 t DM/halyear. There were no residual effects of timing of initial
defoliation in Year 1 , but there was a significant interaction (P<O.O 1 ) defoliation height
Year 1 x defoliation height Year 2 (Table 5 -6). Plots defoliated at 4 cm in Year 1 had
higher production than plots defoliated at 8 cm, and plots intensively defoliated at 4 cm
over 2 years were more productive than the other combinations tested. There were no
differences between treatments defoliated in the 4 cm-8 cm and 8 cm-4 cm sequences
respectively (Table 5 -6).
There were significant differences (P<O.O l ) between introduced cultivars, Grasslands
Goldie being more productive than Steadfast (2.0 and 1 .7 t DM/halyear respectively,
SEM 0.06). Also a significant interaction (P<O.O 1 ) defoliation height Year 1 x
defoliation height Year 2 was observed (Table 5 -6). The same tendency was observed
Chapter Five 1 1 8
for local cultivars, plots defoliated intensively in Year 1 or in both years were more
productive than the other treatments (Table 5-6). There were no residual effects from
timing of defoliation in Year 1 or other interaction effects.
Table 5-6. The effect of defoliation intensity applied in Year 1 and 2 on annual
herbage production (t DM/ha/year) in Year 2 of two groups of BFT cultivars (local
and introduced).
Year 1
4 em
8 em
Defoliation intensity Local cultivars Introduced cultivars
Year 2 (t DM/halyear) (t DM/halyear)
4 em 5 .3 2 .4
8 em 4.9 1 .9
4 em 4.6 1 .8
8 em 4. 1 1 .5
SEM (n) 0 . 1 2 (24) 0.08 (24) Significance * * **
** , P<O.O 1 ; SEM, standard error of the mean; (n) number of observations for each treatment mean
Plots were not defoliated from autumn to early spring, but cultivars differed in autumn
winter activity. The first cut in spring of Year 2 included all accumulated production
from the previous autumn, herbage production being significantly higher for local
cultivars than introduced cultivars (3 . 1 and 0.4 t DMlha respectively, SEM 0.05,
P<O.O 1 ). INIA Draco and San Gabriel produced 65 and 68% of annual production from
autumn to spring, respectively. On the contrary, introduced cultivars Grasslands Goldie
and Steadfast only produced 23 and 29% of annual production in the same period.
5.4.4 Growth rates
The growth rates were analysed for each year during the active growth periods (spring
and summer), under defoliation intervals of 40 days. Contrasting growth patterns
determined the management in the two groups of cultivars (local and introduced)
Chapter Five 1 19
previously described. Differences in timing of defoliation and length of evaluation
periods made it necessary to analyse the information at three levels: individually by
cultivar, by pair of cultivars and by the number of regrowth periods, particularly in Year
1 .
Year 1
First, a preliminary analysis of the general growth pattern by cultivar (Figure 5-3 a-d)
showed maximum growth rates in early summer declining over the season (P<O.OO 1 in
all cases). In the majority of periods, there were higher growth rates on plots when cut
at 4 cm than at 8 cm height. BFT Steadfast and Grasslands Goldie had a shorter and less
productive growth season than INIA Draco and San Gabriel. The growth rates for
Grasslands Goldie and Steadfast in early autumn were poor, less than 1 0 kg DM/ha/day
(Figure 5 -3 c, d), contrasting with rates over 20 kg DM/ha/day for San Gabriel and
INIA Draco (Figure 5 -3 a, b).
There were no differences in growth rate between local cultivars, maximum growth rate
being achieved (P<O.O l ) in December-January, and the more intense defoliation (4 cm)
increased growth rate (P<O.O l ) (Figure 5-4). For introduced cultivars, there was a
significant interaction cultivar x height of defoliation at the last cutting (P<O.O l ),
Grasslands Goldie producing more at 4 cm than at 8 cm height or Steadfast at 4 cm, and
these higher than Steadfast at 8 cm (Figure 5 -5) . However, the low growth rates
measured at this time minimise the importance of these differences.
Chapter Five
60
40
20
60
40
20
--.- 4 cm
___ 8 cm
4 N ov-15 Dee
(t)Studfut
� 4 cm
___ 8 em
NS
1 5 Dec·25Jun
NS
I
25 lan-S M arch 8 M arch· I S A p r
Periods o f growth
NS I
I
4 Dec- I J Jan 1 3 18n-22 Feh 21 feo-S A p r
Periods o f growth
60
40
20
60
40
lO
(b) San Gabrltl
I
-.-4 cm
____ 8 cm
4 N o v · 1 5 Dee
120
N S
I
15 Oee-2S Jan 25 Jan·8 M arch 8 M arch- I S Apr
Periods of growth
(d) GrulllQds Goldit
I NS
I
-.-4 cm
__ 8 em
4 Dec·JJ Jan 1 3 J80-22 Feb U Feb-S A p r
Periods of growth
Figure 5-3. Seasonal variation in growth rates (kg DM/ha/day) of BFT cultivars (a) INIA Draco, (b) San Gabriel, (c) Steadfast and (d) Grasslands Goldie under two defoliation intensities. Vertical bars represent SEM; * * , (P<O.Ol); *, (P<0.05) and NS, (differences not significant) for corresponding growth periods. The number of observations (n) for each treatment mean was 4 (4 Nov-15 Dec), 8 (15 Dec-25 Jan),12 (25 Jan-8 March) and 12 (8 March-IS Apr) for INIA Draco and San Gabrie1. For Steadfast and Grasslands Goldie (n) was 4 (4 Dec-13 Jan), 8 (13 Jan-22 Feb) and 12 (22 Feb-5 Apr).
Chapter Five
SEM
Significance
60
o
1 .68 ...
4 Nov- 1 5 Dec
2.36 NS
1 .60 ... 0.86 ...
I S Dec-25 Jan 25 Jan- 8 March 8 March - 1 5 Apr
. 4 cm 0 8 cm
121
Figure 5-4. Effect of height of defoliation on growth rates of local BFT cultivars from November 1998 to April 1999. (**, P<O.OI ; NS, not significant; SE M, standard error of the mean, n = 8 (4 Nov-1 5 Dec), n=16 (15 Dec-25 Jan), n=24 (25 Jan-8 March and 8 March-15 Apr» .
SEM 3.7 2.2 8.7 S ignificance NS NS ...
60 I�
� 40 l � I
� I Cl Ji 20
0 " "_--J
4 Dec- 1 3 Jan 13 Jan-22 Feb 22 Feb-5 A p r
O Grass lands Gold ie-4 c m . Grass lands Go ld ie-8 c m
/!!J Stead fas t-4 c m f2I Steadfast-8 cm
Figure 5-5. Growth rate of introduced BFT cultivars defoliated at two different heights from December 1998 to April 1999. (**, P<O.OI ; NS, not significant; SEM, standard error of the mean, n=4 (4 Dec-13 Jan), n=8 (13 Jan-22 Feb), n=12 (22 Feb-5 Apr» .
Chapter Five
(a) SEM
Significance
60
.g 40 � ..c:
� Ji 20
o
2 .36 ••
4 Nov- 1 5 Oec 1 5 Oec-25 Jan
2
1 .96 • •
25 Jan-8 Mar
2 3
Number o fregrowths
(b) SEM 1 .58
NS Significance
4 Oec- 1 3 Jan 1 3 Jan-22 Feb 60 I
{ 40 r
1 .05 . .
8 Mar- I S Apr
2 3 4
0.40 • •
2 2 Feb-I S Apr
� I �
�
2
: l�I_� I _ _ ._D� _ _ 2 2 3
Number of re growths
122
Figure 5-6. Growth rates of (a) local cultivars and (b) introduced cultivars of BFT, influenced by the number of previous defoliation during spring and summer. (**, P<O.OI ; NS, differences not significant for corresponding growth periods; SEM, standard error of the mean, n=16). Bars with same colour represent the same treatment across periods.
Chapter Five 123
Because of different times of initial defoliation, there were plots with different numbers
of cuts at the same time. The effect of number of cuts on regrowth is presented in
Figures 5-6 a and b for local and introduced cultivars, respectively. Between local
cu!tivars, there were significant differences (P<O.O I in all cases) in regrowth rates when
the number of cuts increased from one to two, one to three or two to four regrowths
(Figure 5-6 a). The number of successive defoliations affected re growth rates of
introduced cu!tivars, only for the period February-April (Figure 5-6 b), first re growth
being more productive than the second or third re growth (P<O.O l ).
Year 2
The dry conditions of Year 2 limited the growth rates of all cultivars during spring
summer. There were significant differences (P<O.O l ) between local cultivars in the
overall growth rate, in favour of INIA Draco. Also, these materials showed higher
growth rates (P<O.O 1 ) when defoliated at 4 cm rather than 8 cm. There was a residual
effect of the timing of first defoliation (P<O.05) in Year 1 , with high growth rates in
plots defoliated latest in Year 1 (Table 5-7). There were no interaction effects.
There were significant differences (P<O.O l ) between introduced cultivars, Grasslands
Goldie having the best performance. As occurred with local cultivars, defoliation
intensity in Year 2 affected (P<O.O l ) growth rate of introduced materials, 4 cm being
better than 8 cm height (Table 5-7).
The analysis of the different growth periods showed in general that the best growth
conditions occurred from October to December, when the intensity of drought was not
extreme. During summer, the growth rates declined and there were no significant
differences between main treatments.
Chapter Five 124
Table 5-7. Effects of Year 1 and Year 2 treatments on overall growth rates (kg
DM/ha/day) during spring-summer of Year 2 in two pairs of BFT cultivars.
Local c ultivars
Cultivars INIA Draco San Gabriel SEM (n) S ignificance
Defoliation intensity (Year 2) 4 cm 8 cm SEM (n) Significance
Timing effect (residual effect Year /) November December January SEM (n) Significance
1 0.7 9.4
0.34 (48) * *
1 1 .9 8.2
0.34 (48) * *
9.4 9.8
1 1 .0 (32) 0.42
*
Introduced cultivars
Cultivars Grasslands Goldie Steadfast SEM (n) Significance
Defoliation intensity (Year 2) 4 cm 8 cm SEM (n) Significance
1 0.7 7.9
0.42 (48) * *
10.6 8.0
0.42 (48) * *
* * , P<O.O l ; * , P<O.05; SEM, standard error o f the mean; (n) number o f observations for each treatment mean
5.4.5 Sward structure characteristics
Sward characteristics were described at two levels. Firstly, the residual forage after
cutting (section 5 .4 .5 . 1 ) , including the residual herbage mass, leaf area and leaf:stem
ratio of defoliated swards. Also, a description of plant architecture, where a point
quadrat technique was used to describe the tissues distribution of BFT in the sward, is
provided in section 5 .4 .5 .2 .
5.4.5.1 Residual forage.
Local cultivars did not show differences in residual herbage mass, residual leaf area for
regrowth or leaf/(leaf+stem) ratio (Table 5-8). But the height of defoliation determined
a lower residual herbage mass (P<O.O l ) and lower LAI (P<O.O l ) when swards were
defoliated at 4 cm than 8 cm height. Defoliation height did not affect residual
Chapter Five 125
leaf/(leaf+stem) ratio. There were no differences in the parameters when cutting was
done at different growth stages.
Table 5-8. Residual herbage mass, residual leaf area index (LAI) and
leaf/(leaf+stem) ratio after cutting of local cultivars managed at two defoliation
heights and three times of initial defoliation.
Residual herbage mass Residual LAI Residual Leaf/(leaf+stem) Variables (DM kg/ha) (cm2 leaves/cm1 soil) Ratio
Cultivars INIA Draco 920 0.09 0.09 San Gabriel 980 0.07 0.08 SEM (n) 5 1 (24) 0.0 1 0 (24) 0.0 1 0 (24) Significance NS NS NS
Defoliation height 4 cm 660 0.04 0.08 8 cm 1 240 0. 1 3 0.09 SEM (n) 5 1 (24) 0 .0 1 0 (24) 0.0 1 0 (24) Significance ** * * NS
Time of initial cutting Vegetative 920 0.08 0. 1 0 50% Flowering 930 0.08 0.09 Advanced maturity 1 020 0.007 0.08 SEM (n) 58 ( 1 6) 0.032 ( 1 6) 0 .0 1 1 ( 1 6) Significance NS NS NS * *, P<O.O l ; *, P<0.05; NS, not significant; SEM, standard error of the mean; (n) number of observations for each treatment mean
There were no differences in residual herbage mass between introduced cu!tivars, but
Steadfast had a higher residual leaf area index (P<O.05) than Grasslands Goldie (Table
5-9). The height of defoliation significantly affected the residual herbage mass (P<O.O 1 )
and the residual leaf area index (P<O.O l ), 8 cm being higher than 4 cm height in both
cases, and the residual ratio leaf/(leaf+stem) (P<O.Ol ), 4 cm being higher than 8 cm
height.
Chapter Five 126
Table 5-9. Residual herbage mass, residual leaf area index (LA I) and
leaf/(Ieaf+stem) ratio of introduced cultivars managed at two defoliation heights
and three times of initial defoliation.
Variables
Cultivars Grasslands Goldie Steadfast SEM (n) Significance
Defoliation height 4 cm 8 cm SEM (n) Significance
Time of initial cutting Vegetative 50% Flowering Advanced maturity SEM (n) Significance
Residual herbage mass (DM kglba)
1 190 1 240
59 (24) NS
1 0 1 0 1 420
59 (24) **
1 200 1 2 1 0 1 270
54 ( 1 6) NS
Residual LAI (cm2 leaves/cm2 soil)
0.25 0.35
0.035 (24) *
0. 1 6 0.43
0.035 (24) **
0.36 0.27 0. 1 9
0.032 ( 1 6) * *
Residual Leaf/(leaf+stem) Ratio
0. 1 9 0.2 1
0 .024 (24) NS
0. 1 4 0.26
0.022 (24) * *
0.3 1 0 . 1 9
0.0. 1 0 0.024 ( 1 6)
**
* *, P<O.O l ; * , P<0.05; NS, not significant; SEM, standard error of the mean; (n) number of observations for each treatment mean
5.4.5.2 Plant architecture
The vertical distribution of plant components in the sward strata at December 1 999 is
presented in Figures 5-7 and 5-8, for local and introduced cultivars, respectively. A
higher proportion of dead material was distributed between 0 and 1 2 cm height for
introduced cultivars in comparison with local ones. Annual grasses ( Vulpia australis
and Gaudinia fragilis) represented the main components of the dead fraction.
Local cultivars tended to be taller than introduced cultivars, independently of
defoliation intensity applied. A higher proportion of dead stems was observed in local
cultivars than in introduced cultivars between 0- 1 2 cm height, the proportion being
higher for the 8 cm defoliation height treatment. Live stems were better distributed
across strata in local cultivars, and for introduced cultivars stems were mainly in the
first 20 cm strata.
Chapter Five 1 27
The proportion of leaves between 0-4 cm was low in all cases, and between 4-8 cm only
Steadfast defoliated at 4 cm height showed a high frequency of leaves (Figure 5-8 c).
The pattern of leaf distribution in Steadfast changed with the intensity of defoliation,
intensive defoliation increasing the distribution of leaves in low strata.
Reproductive structures (flowers and pods) were placed from 20 cm approximately to
the top of sward, introduced cultivars having a high proportion of flowers than local
cultivars that were mainly in pod stage.
56-60
48-52
:[ 40-44 '" 32-36 � Il 24-28
� 1 6-20
8- 1 2
a) INIA Draco - 4 cm
0-4
----r-'
56-60
48-52
:[ 40-44
� 32-3 6 �
.; 24-28
� 1 6-20
8- 1 2
0-4
o
20
20 40 60 80 l OO 1 20
Nwnber of hi IS per 4 an strmum
b) INIA Draco · 8 cm
40 60 80 l OO 1 20
Number of hits per 4 cm stratwn
56-60
48-52
:[ 40-44
� 32-36 �
; 24-28
� 1 6-20
8- 1 2
c ) San Glbriel - 4 cm . IlFT Lcaf
O B F T Slcm
• BFT F h �'Cr
. I l F T P o d
. BF"!' Dead !Item
. IlFT Dead Ie.r
• Other dead malcnal
0-4 /L-�---
56-60
48-52
:[ 40-44
� 32-36 E ;; 24-28
1 �, 1 6-20
8- 1 2
20 40 60 80 l OO 1 20
Nwnber of hi Is per 4 cm Slratwn
d) San Glbriel • 8 cm
0.4 1---
o 20 40 60 80
Number of hits per 4 cm stratum
lOO 1 20
Figure 5-7. Vertical distribution of tissues in BFT swards strata for local cultivars
under two defoliation heights in December 1 999, determ ined from inclined point
quadrat contacts.
Chapter Five
a) Grasslands Goldie - 4 cm
56-60
48-52
.&0-44
32-36
24-28
16-20
8-12
0-'
20 '0 60 80
Numbcrofhiu per" cm sHlturn
b) (hsslands Goldie - 8 cm
56-60
48-H
32·)6
24-28
16-20
K- 12
0-' ------20 .0 60 .0
Numbcrofus pe r " cmllnllum
100 1 20
100 1 2 0
56-60
48-52
40--44
]'2-36
24·28
16·20
8·12
0-'
c) SleadfaSl - 4 cm
.1-......
- -20 .0 60
a BFTLe. r
O BFT S lcm
. I)FT Fb�r
• 11FT Pod
• 11FT Dead Slem
• BFT Dud IClr
• Qlhcrdcad matcr.l
80 100 1 20
Nwmcr orb .. per" c"'lrll�
d) SleadfaSl - 8 cm
48-52
40-44
32·]6
24-28
16-20
8 · 1 2
I I. .1 •• . : -
---0-' -- ---"" 20 .0 60 '0
Nurnbcroflull pct"4 a.llnlllIIII
100 120
128
Figure 5-8. Vertical distribution of tissues in BFT swards strata for introduced
cultivars under two defoliation heights in December 1 999, determined from
inclined point quadrat contacts.
5.4.6 Nutritive value
Nutritive value (in vitro organic matter digestibility, nitrogen and acid detergent fibre)
of BFT cultivars was compared at the different dates of first harvest and reported as the
nutritive value of accumulated BFT forage. Also, nutritive value was monitored at 40
days defoliation intervals, after first cut in November or December for local and
introduced cultivars respectively. Because of differences in growth patterns, local and
introduced cultivars were compared independently_ A significant interaction effect time
Chapter Five 1 29
x cultivar x defoliation height was observed in some of tested parameters, thus
information presented in Tables 5- 1 0 to 5- 1 3 shows this interaction for all cases and
when it is not significant, the time effect is also shown.
Also in this section is reported the condensed tannin content of cultivars during the
spring of Year 2, when cultivars were at the vegetative stage.
5.4.6.1 Nutritive value of accumulated BFT forage
Local cultivars
A significant interaction time x cultivar x defoliation height was observed in OMD
(Table 5- 1 0). In November, Draco had higher OMD than San Gabriel, independently of
defoliation height, but in December San Gabriel defoliated at 8 cm showed higher
digestibility than the others. In January and after 262 days of herbage accumulation,
there was a general decline in forage quality and no differences between treatments
(Table 5- 1 0) .
There was only a time effect (P<O.O l ) for the nitrogen content of local cultivars,
nitrogen declining over time in all cases (Table 5- 1 0) . The ADF fraction showed
differences over time (P<O.O 1 ), increasing when the accumulation period increased.
There were no differences between cultivars or defoliation heights studied (Table 5 - 10) .
Chapter Five 130
Table 5-10. Nutritive value of local BFT cultivars for the first harvest from sowing
to three different periods of dry matter accumulation for plots defoliated at two
different h eights.
In vitro OMD (glkg) INIA Draco - 4 cm INIA Draco - 8 cm San Gabriel - 4 cm San Gabriel - 8 cm SEM (n) Significance
Nitrogen (g/kg) INIA Draco - 4 cm INIA Draco - 8 cm San Gabriel - 4 cm San Gabriel - 8 cm SEM (4) Significance
ADF (g/kg) lNIA Draco - 4 cm INIA Draco - 8 cm San GabrieI - 4 cm San Gabriel - 8 cm SEM (4) Significance
Sampling dates/ (length of accumulation from sowing)
November 4 (180 days)
663 670 626 647
2.8 (4) *
39 39 30 33
0.1 (4) NS
267 2 1 8 291 2 1 9
5 .3 (4) NS
December 1 5 (22 1 days)
625 65 1 6 1 9 683
4.3 (4) **
27 29 25 28
0.5 (4) NS
328 228 3 1 0 227
5.7 (4) NS
January 25 (262 days)
538 562 495 5 1 5
4.8 (4) NS
1 9 23 1 9 23
0.9 (4) NS
432 383 444 4 1 2
1 2.4 (4) NS
SEM (time*cultivar*height ) [SEM (time)]
1 0.8 **
0 .9 NS [0.6 * * ]
1 l .2 N S [8.4 * * ]
OMD, organic matter digestibility; ADF, acid detergent fibre; SEM, standard error of the mean; (n) number of observations for each treatment mean; * * , P<O.O 1 ; *, P<O.05; NS, not significant; [SEM(time)], presented when SEM (time*cultivar*height) not significant.
Introduced cultivars
A significant interaction time x cultivar x defoliation height (P<O.O l ) was observed for
OMD (Table 5 - 1 1 ). In early December, Steadfast defoliated at 4 cm showed the highest
OMD, but the lowest if defoliated at 8 cm. Grasslands Goldie did not show differences
in OMD as a consequence of defoliation height. In January, there were no differences
between treatments, but Grasslands Goldie showed higher OMD than Steadfast in
February, and each cultivar had higher OMD if defoliated at 8 cm than 4 cm.
Chapter Five 131
Table 5-1 1 . Nutritive value of introduced BFT cultivars for the first harvest from
sowing to three different periods of dry matter accumulation for plots defoliated at
two different heights.
Sampling dates/(length of accumulation from sowing)
December 4 January 1 3 February 22 SEM
(time*cultivar*height) (2 1 0 days) (250 days) (290 days) [SEM {time}]
In vitro OMD (g/kg) Grasslands Goldie - 4 cm 663 6 1 5 52 1 Grasslands Goldie - 8 cm 666 642 542 1 1 .5 * * Steadfast - 4 cm 680 549 484 Steadfast - 8 cm 645 563 5 1 3 SEM (n) 3 .6 (4) 6.3 (4) 1 .4 (4) Significance * * NS *
Nitrogen (g/kg) Grasslands Goldie - 4 cm 33 25 24 Grasslands Goldie - 8 cm 39 28 26 1 .7 ** Steadfast - 4 cm 37 24 23 Steadfast - 8 cm 35 25 24 SEM (n) 0.6 (4) 0.7 (4) 0.2 (4) Significance * * NS * *
ADF (g/kg) Grasslands Goldie - 4 cm 295 379 457 Grasslands Goldie - 8 cm 2 1 9 322 377 8.0 NS Steadfast - 4 cm 2 1 3 403 472 [4.3 * *] Steadfast - 8 cm 224 4 1 1 448 SEM (n) 3 . 1 (4) 5 .8 (4) 3 .3 (4) Significance * * * * * *
OMD, organic matter digestibility; ADF, acid detergent fibre; SEM, standard error o f the mean; (n) number of observations for each treatment mean; ** , P<O.O l ; *, P<O.05; NS, not significant; [SEM(time)], presented when SEM (time*cultivar*height) not significant.
There was a significant interaction time x cultivar x height of defoliation for the
nitrogen content, the decline being significant (P<O.O l ) with the more extended period
of accumulation. The ADF fraction increased over time (P<O.O l ). In December,
Grasslands Goldie defoliated at 4 cm had highest ADF content, but in January the ADF
of Steadfast was higher than in Grasslands Goldie. For the final date, Steadfast
defoliated at 4 cm had the highest proportion of fibre, Grasslands Goldie defoliated at 4
cm and Steadfast defoliated at 8 cm were intermediate, and Grasslands Goldie cut at 8
cm lowest (Table 5 - 1 1 ).
Chapter Five 132
5.4.6.2 Nutritive value parameters under regular defoliation
For local cultivars, there were four periods of re growth from November to April during
Year 1 (see Table 5 -2). A significant interaction time x cultivar x defoliation height
(P<O.O l in all cases) was observed for OMD, nitrogen and ADF content (Table 5 - 1 2).
The OMD of San GabrieI was higher (P<O.O 1) than INIA Draco for the first period, but
at the end INIA Draco defoliated at 8 cm height had higher OMD than San Gabriel
defoliated at 4 or 8 cm and INIA Draco at 4 cm. There was a tendency for OMD to be
lower in January and March sampling and high in December or April (Table 5 - 1 2).
The content of nitrogen increased during the last sampling (April), with no differences
between cultivars or defoliation heights. Differences between cultivars were detected in
December, San Gabriel having higher content than INIA Draco independently of
defoliation intensity (Table 5 - 1 2).
Levels of ADF for San Gabriel increased from December to January, and then were
maintained over time (Table 5- 1 2). However, INIA Draco increased ADF levels from
December to March, but reduced ADF in the April sampling. Comparatively, INIA
Draco had higher ADF concentration, except in April where San Gabriel was higher for
the average of defoliation heights than INIA Draco. Differences in ADF concentration
caused by defoliation intensity were clearly detected in April , 4 cm being higher than 8
cm independently of cultivars.
Chapter Five 133
Table 5-12. Nutritive value of local BFT cultivars managed at two defoliation
heights over 40 day intervals from November 1998 to April 1999.
Periods of accumulation SEM
4 Nov-I S Dcc 1 5 Dec-25 Jan 25 Jan-8 Mar 8 Mar- I S Apr (time*cultivar*height )
In vitro OMD (g/kg) INIA Draco - 4 cm 638 589 585 647 INIA Draco - 8 cm 6 1 8 6 14 640 7 1 7 1 8 .5 * * San Gabriel - 4 cm 670 639 639 677 San Gabriel - 8 cm 662 622 626 695 SEM (n) 5.7 (4) 7.0 (4) 2.4 (4) 8.2 (4) Significance * * NS ** **
Nitrogen (g/kg) INIA Draco - 4 cm 28 25 27 35 INIA Draco - 8 cm 25 25 29 3 8 1 .5 * * San Gabriel - 4 cm 3 1 27 29 37 San GabrieI - 8 cm 30 26 27 3 7 SEM (n) 0.2 (4) 0.6 (4) 0.5 (4) 0 . 1 (4) Significance * * NS ** NS
ADF (g/kg) INIA Draco - 4 cm 3 1 5 365 405 338 INIA Draco - 8 cm 3 1 0 367 348 249 1 7.8 * . San Gabriel - 4 cm 258 3 1 3 322 322 San Gabriel - 8 cm 230 328 309 29 1 SEM (n) 5 . 1 (4) 8. 1 (4) 6.3 (4) 9.0 (4) Significance * * ** * * * *
OMD, organic matter digestibility; ADF, acid detergent fibre; SEM, standard error o f the mean; (n) number of observations for each treatment mean; * *, P<O.O l , NS; not significant.
For introduced cu!tivars, there were three periods of re growth from December to April
of Year 1 (see Table 5-2). A significant interaction time x cultivar x defoliation height
was observed for OMD (P<O.O I ) and ADF content (P<O.O I ) (Table 5 - 1 3) .
OMD increased over time for Steadfast (4 and 8 cm height) and Grasslands Goldie- 4
cm, but Grasslands Goldie-8 cm showed higher OMD than others in January, mantained
its value in February and then declined in April . In January, there were large differences
between cultivars, Steadfast having lower OMD than Grasslands Goldie. The OMD of
Grasslands Goldie-8 cm fel l sharply from February to April .
Chapter Five 134
The ADF content of Grasslands Goldie did not differ between sampling dates, but for
Steadfast-4 cm concentration was highest in February and for Steadfast-8 cm in January
(Table 5 - 1 3) .
Table 5-13. Nutritive value of introduced BFT cultivars managed at two defoliation
height and 40 day intervals from December 1 998 to April 1999.
In vitro OMD (glkg) Grasslands Goldie - 4 cm Grasslands Goldie - 8 cm Steadfast - 4 cm Steadfast - 8 cm SEM (n) Significance
Nitrogen (g/kg) Grasslands Goldie - 4 cm Grasslands Goldie - 8 cm Steadfast - 4 cm Steadfast - 8 cm SEM (n) Significance
ADF (g/kg) Grasslands Goldie - 4 cm Grasslands Goldie - 8 cm Steadfast - 4 cm Steadfast - 8 cm SEM (n) Significance
Periods of accumulation
4 Dec- 1 3 Jan 1 3 Jan-22 Feb 22 Feb-5 Apr
643 677 675 660 660 6 1 9 626 653 67 1 605 665 657
2.6 (4) 5 .4 (4) 5.7 (4) ** NS * *
28 36 3 7 2 8 3 2 3 2 3 0 36 39 32 35 36
0.7 (4) 0.4 (4) 0.5 (4) ** ** * *
3 1 9 323 3 1 7 330 346 325 297 342 304 336 305 295
5 .5 (4) 6.7 (4) 7.6 (4) * * * * NS
SEM (time*cultivar*height )
[SEM (time»)
2 1 .4 * *
0.4 N S [0.4 * * ]
22. 1 **
OMD, organic matter digestibility; ADF, acid detergent fibre; SEM, standard error of the mean; (n) number of observations for each treatment mean; ** , P<O.O l ; NS, not significant; [SEM(time)], presented when SEM (time*cultivar*height) not significant
The nitrogen content increased over time (P<O.O l ), Steadfast having higher content
than Grasslands Goldie in January. Grasslands Goldie defoliated at 8 cm tended to have
lower nitrogen content during February and April than the other treatments. When
significant, differences by defoliation height tended to show higher nitrogen
concentration in plots defoliated at 4 cm than at 8 cm. Steadfast-8 cm had higher
nitrogen concentration than Steadfast-4 cm only in January (Table 5 - 1 3) .
Chapter Five 135
5.4.6.3 Condensed tannin content
A preliminary analysis of tannin content was done comparing the cultivars under study
at the vegetative stage in September 1 999. Total CT concentrations showed differences
(P<0.05) between cultivars, San Gabriel having the highest content, Grasslands Goldie
and INIA Draco intermediate and Steadfast the lowest. Over 86% of CT was bound
(Table 5 - 14), with the largest component being protein-bound. San Gabriel showed a
tendency (P=0.097) to have a higher protein-bound fraction than Steadfast.
Table 5-14. Condensed tannin (CT, g/kg DM) contents of four BFT cultivars at
vegetative stage in early spring.
Cultivars Extractable Protein-bound Fibre-bound Total Bound CT CT CT CT CT (% total)
San Gabriel 4.2 25 . 1 1 .4 30.7 86.3 INIA Draco 4.3 20.7 0.9 25.9 83.4 Grasslands Goldie 2.5 1 9.7 1 .8 24.0 89.6 Steadfast 3 .4 1 6.8 1 .4 2 1 .6 84.2 SEM (n) 0.58 (2) 1 .48 (2) N/A 0.97 (2) Significance NS (0.097) '"
"', P<0.05; NS, not significant; N/A, not replicates for statistical analysis; SEM, standard error of the mean; (n) number of observations for each treatment mean
5.4.7 Plant density
Plant density was evaluated at four stages of the experiment from July 1 998 to March
2000. The initial sowing rate was 875 viable seeds/m2. After 1 1 0 days from sowing
(July 1 998), overall plant density was 273 plants/m2, increasing to 409 plants/m2 in
March 2000, as a consequence of further late germination during spring. The
establishment showed significant differences (P<O.O I ), San Gabriel having the highest
density and Grasslands Goldie the lowest (Figure 5-9). Over the fol lowing sampling
dates, significant differences between cultivars (P<O.O I in all cases) were observed, San
Gabriel and Steadfast always having the denser stands. During the second year, there
was a general decline in density in all cultivars. In October 1 999, plant density was 1 8%
of that measured in March 1 999, and in March 2000 the density recorded was only 44%
of the plants surviving in October 1 999. The final plant density was 3 3 plants/m2. There
Chapter Five 136
was a significant time x cultivar effect (P<O.O l ) on the average of the plant density over
the period (Figure 5-9), because the increase in population from July 1 998-March 1 999
by late germination after July 1 998 was higher in Steadfast than others. A higher
proportion of hard seeds was found in Steadfast than the others in germination tests
done previously.
600
500
400 Na
-..... r/l 300 -I: oS
0..
200
100
0 Sowing (8/5/98)
-0- INIA Draco
-)t( - San Gabriel (29.6) -fr- Grasslands Goldie
•• _ Steadfast
July 1 998 March 1 999 October 1 999 March 2000
Dates of sampling
Figure 5-9. Changes in plant density of four BFT cultivars over the period July
1 998 - March 2000. Numbers in brackets indicate the SEM between treatments at
corresponding dates, (n=24).
5.4.8 Plant morphology
Firstly, individual plant morphology parameters were analysed at an early stage of plant
development ( 1 1 0 days after sowing) to determine characteristics of initial growth of
cultivars. After that, plant morphology parameters at the end of each cutting season
were analysed individually (anova analysis) and integrated (canonical discriminant
analysis). Parameters tested included the number of primary, secondary and new
Chapter Five 137
developing shoots per plant, root diameter, crown and root mass, and total aboveground
mass per plant.
Initial development
BFT San Gabriel was the cultivar that significantly exhibited (P<O.OS) an early
development in plant height, fol lowed by INIA Draco and Steadfast, and Grasslands
Goldie the shortest (Plate 5-2 a). Also, San Gabriel had (P<O.OS) the highest
accumulation of biomass above ground, Grasslands Goldie and Steadfast having the
lowest accumulation (Table 5 - 1 5). There were no differences in root length, stem
number and below-ground mass per plant at this stage of development. However, the
abovelbelow-ground ratio was higher for local cultivars than introduced ones (Table S
I S). San GabrieI had the high ratio (3.9) and Grasslands Goldie the lowest ( 1 .9).
Table 5-15. Morphological parameters of four BFT cultivars 1 1 0 days after a
sowing made on May 8, 1998.
Cultivars Plant height Root length Primary stems Above-ground mass Below-ground mass !cm� !cm� !no.!l!lant� �g/l!lant� !gLl!lant�
San Gabriel 8 6 4 0.089 0.023 INIA Draco 5 6 4 0.062 0.020 Grasslands Goldie 3 6 5 0.035 0.0 1 8 Steadfast 5 6 5 0.044 0.020
SEM (n) 0.4 (24) 0.4 (24) 0.3 (24) 0.0056 (24) 0.0022 (24) Significance * NS NS * NS
* , P<0.05; NS, not significant; SEM, standard error of the mean; (n) number of observations for each treatment mean
The development of rhizomes was explored in the cultivar Steadfast (Plate S-2 b).
During the first year, there were no plants with rhizomes. Rhizome development was
observed in 1 1 % of a sample of 240 plants in April 1 999, increasing to 1 7% in October
1 999. Well-developed rhizomes exceeded SO mm in length, the maximum number
being 3 per plant.
Chapter Five 138
Plate 5-2. BFT plants (6 months old) of San Gabriel and Grasslands Goldie (a),
and one year old plant of Steadfast showing the development of rhizomes (b).
Chapter Five l39
Morphological parameters under defoliation
A significant interaction cultivar x time x height of defoliation was observed in some of
morphological parameters tested at the end of each of the two years. Tables 5 - 1 6 and 5-
17 show treatment values for each year, with some discussion of main effects,
particularly differences between cultivars.
Year 1
A the end of the first year, the cultivar x time x height of defoliation interaction was
significant (P<O.O 1 ) for root and crown weight (Table 5 - 1 6). There was a strong
influence of cultivars for root weight, Grasslands Goldie showing the greatest root mass
(0. 12 , 0. 1 1 , 0.69 and 0.28 g/plant for INIA Draco, San Gabriel, Grasslands Goldie and
Steadfast respectively; SEM 0.020; P<O.O 1 ).
There were differences in crown weight between cultivars, Grasslands Goldie having
the greatest and Steadfast the lowest crown mass (0.36, 0.24, 0. 1 7 and 0.06 g/plant for
INIA Draco, San Gabriel, Grasslands Goldie and Steadfast respectively; SEM 0.0 1 2;
P<O.O I) .
Root diameter differed only between cultivars, the mean value for Grasslands Goldie
being higher than the other cultivars, and also INIA Draco differed from Steadfast (6, 5 ,
4 , and 4 mm for Grasslands Goldie, INIA Draco, San Gabriel and Steadfast; SEM 0.20;
P<O.O 1 ). Also the number of primary shoots differed significantly (P<O.O 1 ) between
cultivars, being higher in GrassIands GoIdie (7 shoots/plant) than the others cultivars (3
to 4 shoots/plant, SEM 0.020; P<O.O I ).
Chapter Five 140
Table 5-16. Individual plant morphology parameters in BFT cultivars affected by
the intensity and timing of defoliation at the end of the first year (March 1999).
Cultivar x height x timing Root weight Crown weight Root diameter Primary ,hoots Secondary shoots New shools
(glplanl) (glpl •• I) (mm) (noJplant) (no./plant) (noJplant)
INIA Draco 4 1 0. 1 7 0.33 5 4 8 4 4 2 0. 1 2 0.3 1 4 4 5 4 4 3 0. 1 3 0.43 5 4 7 3 8 I 0.09 0.25 4 3 8 3 8 2 0. 1 2 0.54 4 4 8 3 8 3 0.09 0.3 1 4 4 7 3
San Gabriel 4 I 0. 1 2 0.23 4 3 5 2 4 2 0.08 0.2 1 4 3 3 1 4 3 0.09 0.30 4 4 5 2 8 1 0. 1 5 0.2 1 4 3 5 2 8 2 0. 1 3 0.22 4 3 5 2 8 3 0.09 0.30 4 3 4 2
Grasslands Goldie 4 1 0.66 0. 1 7 5 8 1 4 4 4 2 0.88 0.23 7 7 14 2 4 3 0.59 0. 1 1 6 6 1 2 2 8 I 0.58 0. 1 4 5 7 8 2 8 2 0.59 0. 1 8 6 7 1 0 2 8 3 0.83 0.2 1 6 8 1 0 2
Steadfast 4 1 0. 1 8 0.06 3 3 6 2 4 2 0.29 0.07 4 5 7 I 4 3 0.22 0.04 3 3 6 1 8 1 0.23 0.05 3 4 7 1 8 2 0.29 0.06 4 4 7 I 8 3 0.48 0.08 4 4 8 1
SEM 0.07 1 0.029 0.5 0.9 1 .7 0.6 Significance (cultivar x height x timing) * * * * NS NS NS NS
**, P<O.O l ; NS, not significant; SEM, standard error of the mean; (n=4); time ( 1 , 2 and 3), correspond approximately to vegetative, 50% flowering and advanced maturity stages respectively
There was a significant interaction cultivar x defoliation height (P<O.05) for the number
of secondary shoots per plant. Grasslands Goldie reduced the number of shoots/plant
from 1 3 to 9 when defoliated at 8 cm rather than 4 cm, but for other cultivars
shoots/plant remained unchanged with defoliation intensity.
The number of dead secondary shoots was analysed as a measure of potential
differences in the production of new growth sites between cultivars or management
variables. Dead shoots per plant were affected by defoliation height (P<O.O l ), initial
time of defoliation (P<O.O 1 ), and cultivars (P<O.O 1 ). Grasslands Goldie had 6 dead
shoots per plant, higher than the rest of the cu]tivars. Plants that were first defoliated
earlier had more dead shoots that plants defoliated late (5 and 3 shoots per plant
Chapter Five 141
respectively). Also, plants defoliated at 4 cm height had more dead shoots that those
defoliated at 8 cm (4 and 3 shoots per plant respectively).
The number of new developing shoots differed significantly (P<O.O 1 ) between cultivars,
INIA Draco having the highest number of new shoots per plant and Steadfast the lowest
(3 , 2, 2 and 1 for INIA Draco, San Gabriel, Grasslands Goldie and Steadfast
respectively; SEM 0.23).
Year 2
The interaction cultivar x defoliation height x time was significant (P<O.O l ) for root
weight, crown weight and root diameter (Table 5 - 1 7). For root and crown weight, there
were significant differences between cuItivars at the end of the experiment. The root
weight was highest in Grasslands Goldie, and San Gabriel and in INIA Draco lowest
(0.48, 0.20, 0.08 and 0.07 g/plant for Grasslands Goldie, Steadfast, INIA Draco and San
Gabriel respectively; SEM 0.0 1 7; P<O.O l ). But INIA Draco had the biggest crown mass
(0.25, 0. 12 , 0. 1 0 and 0.04 g/plant for INIA Draco, Grasslands Goldie, San Gabriel and
Steadfast respectively; SEM 0.007; P<O.O l) . Also, there were differences between
cultivars in root diameter (4, 3 , 3 and 2 mm for Grasslands Goldie, INIA Draco,
Steadfast and San Gabriel respectively, SEM 0. 1 ; P<O.O l ).
Numbers of primary shoots differed between cultivars (5, 3 , 3 and 1 shoots/plant for
Grasslands Goldie, Steadsfast, INIA Draco and San Gabriel respectively; SEM 0.24;
P<O.O l ). There was a significant interaction cultivar x defoliation height (P<0.05) in the
number of active secondary shoots. Grasslands Goldie declined from 9 to 6 secondary
shoots per plant when defoliated at 8 cm instead of 4 cm, but the other cultivars were
not affected. The number of new shoots differed between varieties, (2 new shoots/plant
for INIA Draco and Grasslands Goldie and I shoot/plant in San Gabriel and Steadsfast;
SEM 0. 1 6; P<O.O I ). Also, there were differences in the number of dead secondary
shoots between cultivars (P<O.O 1 ), and defoliation heights (P<O.O 1 ) . Grasslands Goldie
showed 4 dead secondary shoots/plant, two more than the other cuItivars, and also when
Chapter Five 142
it was defoliated at 4 cm had more than at 8 cm height (3 and 2 shoots/plant
respectively).
Table 5-17. Individual plant morphology parameters in BFT cultivars affected by
the intensity and timing of defoliation at the end of the second year (March 2000).
Cultivar x height x timing Root ""eight Crown weight Root diameter Primary shoots Secondary shoots New sboots
(glpl.n!) (glplan.) (mm) (no.lplant) (noJplanl) (noJplant)
INIA Draco 4 1 0. 1 2 0.23 3 3 5 3 4 2 0.08 0.2 1 3 2 4 3 4 3 0.09 0.30 4 3 5 2 8 1 0.D7 0. 1 8 3 2 6 2 8 2 0.08 0.38 3 3 6 2 8 3 0.D7 0.2 1 3 3 5 2
San Gabriel 4 1 0.05 0.09 2 2 3 I 4 2 0.03 0.09 2 2 2 I 4 3 0.03 0 . 1 2 2 2 3 2 8 1 0.D7 0.09 2 2 4 2 8 2 0.2 1 0.09 2 2 3 2 8 3 0.04 0.09 2 2 3 2
Grasslands Goldie 4 I 0.46 0. 1 2 4 6 1 0 3 4 2 0.62 0 . 16 5 5 1 0 2 4 3 0.4 1 0.D7 4 4 8 1 8 I 0.4 1 0. 1 0 4 5 6 1 8 2 0.42 0. 1 2 4 5 7 1 8 3 0.58 0. 1 4 4 5 7 2
Steadfast 4 1 0. 1 2 0.04 2 2 4 4 2 0.20 0.05 3 4 5 4 3 0. 1 5 0.03 2 2 4 8 I 0. 1 6 0.04 2 3 5 8 2 0.2 1 0.04 3 3 5 8 3 0.33 0.06 3 3 5
SEM 0.04 1 0.0 1 8 0.3 0.6 1 .2 0.4 Significance (cultivar x height x timing) * * * * * * NS NS NS
**, P<O.O I ; NS, not significant; SEM, standard error of the mean; (n=4); time ( I , 2 and 3), correspond approximately to vegetative, 50% flowering and advanced maturity stages respectively applied during 1 999.
The ratio above/below-ground biomass was consistently lower for introduced cultivars
compared with local ones at all stages (average being 0.9- 1 .0 and 1 .2- 1 .4 for introduced
and local cultivars respectively).
5.4.8.1 Integrated morphology analysis
The total canonical structure based on individual plot values (Figure 5 - 1 0 a) explained
97% of the variation for Year 1 and 92% for Year 2 (Figure 5- 1 0 b). For canonical
Chapter Five 143
variate 1 , the root weight and crown weight were the most significant parameters, and
for canonical variate 2 the above-ground mass per plant, root diameter and primary
shoots were major contributors, in both years.
Above mass Root diameter Primary shoots
+ a ) March 1 999 t 3
2 (23%)
0 '" 1l A 11 1::. j 0 r1
I::,. sf � cl � 1::,. 1::. A A - I 0
1::. 0 -2
-3 -3 -2
Crown weight
CUL T I VAR 0 0 0
Above mass Root diameter Primary shoots
+ b ) March 2000
t 3 0
2 0 (38%)
'" � 0 1l 11 0 .� 0 :> 0 0 cl 11 0 - I
-2
I::,.
*
- I o Can. Variate 1 (74%)
0
0 cro O
* *
0
* *
* * * ** ** * � * *
2 3 + . -----I�� Root weight
I N I A Draco 0 0 0 G . Go l d i e II II ll San Gabr i e l * * * Steadfast
0 0 0
o <6 §
0 0 0 0
0
* * *
** * * * * * * * * * * "* '\ 1::. I:.J.
* * I:.J. 1::. A I::,. -3 �-r�-��--r-�-I::,.�-r����--���--�-r�--��-'r-��--r-,
-3 Crown weight
-2 - I o Can. Variate 1
I (55%)
2 3 + . -------I�� Root weight
CUL T I VAR 0 0 0 I N I A Draco 0 0 0 G . Go l d i e A A A San Gabr i e l * * * Steadfast
Figure 5-10. Canonical analysis for morphology parameters in the Year 1 (a) and
Year 2 (b).
Chapter Five 144
Plants of cultivar Grasslands Goldie had strong roots, and more primary shoots than the
other cultivars in Year 1 , and together with Steadfast, had higher root weight than San
Gabriel and INIA Draco. At this stage, there were few contrasts between local cultivars.
In Year 2, the four cultivars appeared more clearly differentiated (Figure 5 - 10 b).
Grasslands Goldie showing high root mass, root diameter, shoots per plant and above
ground mass per plant. Steadfast presented high root mass. But local cultivars appear
more clearly differentiated, San Gabriel showed lower above-ground mass, number of
primary shoots and root diameter than INIA Draco, but both cultivars exhibited lower
root and crown mass than introduced cultivars.
5.5 DISCUSSION
The main results of this experiment are grouped for discussion, including firstly the
productivity and adaptability of cultivars, secondly growth, plant type and defoliation
management, thirdly the quality of BFT forage and finally the option for forage
accumulation ofBFT during late spring and summer.
5.5.1 Productivity and adaptation of BFT to the eastern region of Uruguay
Annual herbage production differed significantly between cultivars, being between 2.5
to 2 .6 times higher for local cultivars than introduced cultivars (Figure 5-2, Table 5 -4),
differences occurring over all seasons. Annual production of local cultivars was
comparable with results reported by Fonnoso ( 1 993) with the cuItivar San Gabriel,
particularly during the first year where environmental conditions were not limiting for
growth. Some differences in production can be attributed to the degree of winter activity
between the two groups of cultivars; traditional cultivars used in Uruguay all being
winter active. The active growth period for introduced cultivars occurred from
November to early April . The importance of autumn-winter contribution to annual
Chapter Five 145
production in this study was overemphasised, because drought conditions determined a
poor spring-summer contribution in Year 2 . The cultivar Steadfast was developed in
Missouri, USA (latitude 3 8° 4 ' N) and Grasslands Goldie in New Zealand (higher than
latitude 36° S), contrasting with local cultivars developed in an area lower than 34° 5 '
latitude. High latitudes have shorter growing season, with longer daylenght and cooler
temperatures (Alison and Hoveland, 1 989) than those in Uruguay.
When winter active cultivars from South America were sown in New Zealand, autumn
growth was associated with latitude, performance increasing when sown in low latitudes
of New Zealand, North Island (Charlton et al. , 1 978; Widupp et al., 1 987). In addition,
frost damage in autumn and spring affected performance. Some results with alfalfa
showed that when winter active cultivars were grazed in winter, spring production was
lower than dormant cultivars, because of a significant reduction in root reserves (White
and Lucas, 1 990), demanding a differential management policy. The reported risks of
inappropriate winter defoliation in lucerne may also apply to BFT, particularly in
environments with cold winters. Also, the avoidance of grazing in early spring allows
plants to restore root reserves, improving spring growth (White and Lucas, 1 990).
Herbage production increased when plots were defoliated at 4 cm instead of 8 cm
height, behaviour observed during the two years. These results agree with those of
Pierre and lackobs ( 1 953) and Duell and Gausman ( 1 957) comparing defoliation
intensities of 2 .5 versus 1 0 cm height and 2 .5 versus 7 .5 cm height respectively. Both
studies reported that the magnitude of differences during the second year in favour of
the 2 .5 cm treatment was lower than during the first year, a consequence of a severe
decline in stand density in all treatments. Pierre and lakcobs ( 1 953) mentioned higher
production from lax defoliation at the third year of evaluation. Work by Nelson and
Smith ( 1 968) and Greub and Wedin ( 1 97 1 ) showed a more vigorous regrowth when this
originated from shoots close to crown. Additionally, it can be suggested that intervals
between successive defoliation (40 days) were appropriate, as was the extended rest
interval between mid-autumn to spring at the end of Year 1 , for plants to replenish
reserves. Intensive defoliation (4 cm) was not so severe as previously reported by
Pierre and lackobs ( 1 953) and Duell and Gausman ( 1 957) or results of Chapter 3 of this
Chapter Five 146
thesis where herbage production increased if more lax defoliation applied (2< 6 < 10 cm
height).
The establishment of Grasslands Goldie was the lowest, the cultivar achieving a
reduced population, that could have incidence on DM and morphological results
reported. The slow establishment of introduced cultivars suggested that earlier sowing
dates could allow a better establishment. On the other hand, early spring sowing could
be preferable to late autumn sowing (1. Bologna, personal communication), because the
low establishment and poor competitive capacity could produce a poor stand. However,
final establishment of plants sown in spring wil l be associated with the absence of
drought conditions in summer (Canimbula et al. , 1 994). In the conditions of the trial,
the competition from annual grasses was higher for introduced than local cultivars
because of prostrate growth form of the former from autumn-mid spring. In Uruguay,
local BFT cultivars are tolerant to sowing dates in late autumn, comparatively more
than white clover (Canimbula et al., 1 994), but an important additional establishment
from late germination seed was observed in this trial in spring (Figure 5 -9).
5.5.2 Growth, plant type and defoliation management
Growth rates were highest between December - January for both local and introduced
cultivars, then declined to autumn (Figure 5-3). Local cultivars registered maximum
rates between 45-50 kg DMIha/day during the first year. These results agreed with those
reported by Formoso ( 1 993), for a series of 1 7 experiments between 1 963- 1 990
conducted at la Estanzuela, Uruguay working with BFT San Gabriel. For introduced
cultivars, maximum growth rates were 2 1 and 28 kg DMIha/day for Grasslands Goldie
and Steadfast respectively during the first year. The extreme drought conditions
registered during the second year (Figure 5 - 1 ) reduced growth rates and plant density
significantly in al l cultivars. This decline can not be attributed to disease incidence,
because only minor symptoms were detected on roots when plant morphology
measurements were done. Also the trial was established on an area of native grasses
without history of BFT or other crops.
Chapter Five 147
The intensity of defoliation affected growth rates, which were always greater for 4 cm
than 8 cm defoliation, as was discussed for total herbage production. Introduced
cultivars showed higher residual leaf area than local ones, in particular Steadfast.
Residual leaf area in local cultivars was extremely low for regrowth, which did not
increase much under lax defoliation (8 cm), re growth being more dependent on plant
reserves than residual leaf area. In addition, there were few leaves in lower strata
(Figures 5-7 and 5-8) because of drought conditions. Tolerance to drought of local
cultivars can be considered higher than introduced ones if productivity is observed, but
in terms of survival of plants the decline in density during the second year was similar
in all cultivars.
These findings suggest that environmental constraints and adaptability rather than post
defoliation residues detennined growth responses of different cultivars. The residual
leaf area for local cultivars did not appear to exert a strong influence, despite the erect
growth habit. The levels of root reserves could be determining these effects, and
defoliation intervals of 40 days applied during spring-summer and permitting rest from
autumn to early spring may have allowed plants to rebuild reserves for successive
regrowths. When defoliation intensity was a significant factor influencing plant
morphology, the number of shoots tended to be higher in plants defoliated hard (4 cm
height), suggesting that plants reacted by providing more new regrowth sites. Regrowth
rates declined with successive defoliation (Figure 5 -6), an effect which can be attributed
to the decline in root reserves levels for plants progressively defoliated.
Introduced cultivars showed a lower ratio of abovelbelow-ground biomass compared
with local cultivars ( 1 .9-2.2 vs. 3 . 1 -3 .8 respectively). High ratios for abovelbelow
ground mass found in local cultivars indicate that they could be sensitive to hard
defoliation, more than introduced cultivars (Table 5 - 1 5) . Although this is an arbitrary
value, a more conservative strategy and reduced risk to plant survival could be
attributed to plants with a low abovelbelow-ground ratio As well , the more prostrate
plant structure of introduced cultivars could have advantages in terms of regrowth from
a higher residual leaf area. Thus, the environmental constraints may have a stronger
Chapter Five 148
influence on the performance of introduced cultivars than the morphological structure
for growth, local cultivars showing characteristics less tolerant to hard defoliation. The
development of INIA Draco cultivar, with improved conditions of crown size, shoots
and leaf density (Rebuffo and Altier, 1 996) was corroborated in the experiment.
However, this did not result in productive advantages during the first two years of the
BFT pasture. It is possible that potential advantages may be expressed later, and the
disease pressure in the area of the trial was not severe enough to test the better degree of
resistance of INIA Draco over San Gabriel.
The presence of rhizomes has been suggested as a useful character to increase BFT
persistence particularly in those environments where diseases pressure is high (Li and
Beuselinck, 1 996). Only a portion of plants of BFT Steadfast expressed this character,
rhizome development occurred in autumn when daylength and temperature declined.
Steadfast does not appear to be a suitable option for Uruguayan environments, but the
potential advantages of the presence of rhizomes could be achieved if this character is
transferred to cultivars with winter activity like San Gabriel or Draco, increasing
persistence without losses in productivity.
5.5.3 Forage quality
In general, the results showed a high nutritive value of herbage of all BFT cultivars
evaluated, in many cases comparable to lucerne, a recognised high quality forage
(Spedding and Diekmahns, 1 972, cited by Frame et aI. , 1 998).
When BFT was defoliated regularly, organic matter digestibility varied from 590 to 700
g/kg for local cultivars and from 600 to 680 glkg for introduced ones. The decline that
is reported in other trials to end of summer (Formoso, 1 993) did not occur in this trial,
and OMD increased from summer to early autumn. The drought that affected the trial
during summer reduced the proportion of leaves in lower strata, reducing OMD, and the
increase in rainfall at the end of first summer promoted new regrowth, increasing OMD.
Chapter Five 149
In other conditions, drought increased digestibility by an improved leaf:stem ratio and
delaying maturity (Petersen et al., 1 992 cited by Frame et al. , 1 998).
Plant nitrogen content was high, reflecting the high nutritive value reported for BFT.
For local cultivars when managed under regular defoliation, nitrogen ranged between
25 and 38 g/kg DM and for introduced cultivars between 30 and 36 g/kg DM. These
results can be compared with those obtained by Acufia ( 1 995) for a series of winter
active cultivars from South America (including San Gabriel) where nitrogen ranged
between 26 and 32 g/kg DM, or by Formoso ( 1 993) for San Gabriel between 29 and 37
g/kg DM. Changes in nitrogen over time were not consistent with those reported by
Formoso ( 1 993), who showed a steady and significant decline over the summer period.
In the current study, drought conditions during the first summer were fol lowed by rain
in early autumn, allowing a regrowth of cultivars and consequently increasing the
nitrogen content of cut forage.
ADF content ranged from 230 to 400 g/kg DM in local cultivars and from 290 to 350
g/kg DM in introduced cultivars. Acufia ( 1 995) reported variations between 240 and
380 glkg DM for a group of South American accessions.
Condensed tannins results (Table 5 - 14) need to be carefully interpreted because of the
l imited number of samples involved in the analyses. The total CT content varied from
22 to 3 1 glkg DM, concentrations within the recommended range of 20-40 g/kg DM
that produce beneficial effects on bloat protection, intake levels and fibre digestibility
compared with higher concentrations in the diet (Barry, 1 989; Waghom et al., 1 990).
San Gabriel showed the highest content, the bound-protein fraction being the largest
component.
5.5.4 Forage accumulation
The availability of quality forage in summer is crucial for many systems, especially for
young animal categories that demand good forage quality. In conditions of Uruguay, C4
grasses are the main components of native swards, thus the addition of a high quality
Chapter Five 150
protein bank of BFT can play a strategic role. Stockpiling BFT forage to late spring or
summer is possible, but losses in forage mass and nutritive value determine the levels
and periods of accumulation. In this trial, the use of an early spring growth cultivar like
BFT San Gabriel showed advantages for late spring accumulation. But the rapid decline
in ordigestibility in San GabrieI after flowering (Figure 5 - 1 1 a) suggests that INIA
Draco could be recommended as a better cultivar for mid summer (Figure 5 - 1 1 b). In
San Gabriel, periods of accumulation after 50% flowering showed no advantages in the
levels of digestible organic matter accumulated (Figure 5- 1 1 a). In general, stems
decline in organic matter digestibility much faster than leaves do and the increase in
stem proportion at mature stages produces a large influence on total digestibility
(Buxton et al. , 1 985). The low levels of forage accumulation of introduced cultivars
(Table 5-6) in comparison with local cultivars meant that there was no advantage to be
gained in digestible organic matter production from an extended period of growth in
spring. However, these results relate to first year swards, and patterns of accumulation
may be different in older swards.
Chapter Five 151
(a) San Gabriel
4 70
OMD Cl � S2-" � 0 3 60 .� :.0 :; "" .� s '"
::l � 0) ,-., 8 � .::P � '" � '" Cl �
� � " ::: '" '" -e 2 50 E dl <.> '2
MA '" g
Advanced maturity 40 50% Flowering
Vegetative
0
Sowing (8-May) 4-Nov I S-Dec 25-Jan
(b) INIA Draco
4 70
Cl OMD DM � " � 0 3 60 .� :.0 :; � .� S '" O) � " � .� � 8 � "' � '" Cl �
� � � '" -e 2 50 E " DOMA <.> ::r: 'c '" g
40 Advanced maturity
50"10 Flowering
0
Sowing (8-May) 4-Nov IS-Dec 2S-Jan
Figure 5-1 1 . Changes in dry matter harvested (DM), digestible organic matter
harvested (DO MA), and organic matter digestibility (OMD) of two BFT cultivars
(a) San Gabriel and (b) INIA Draco from sowing to three physiological stages.
Chapter Five
5.6 CONCLUSIONS
152
The levels of productivity achieved by tested BFT cultivars in the eastern region of
Uruguay showed that local cultivars are more adapted and productive than
introductions from USA and New Zealand, production being consistently higher over
the evaluation. There were differences in the degree of winter activity of cultivars.
Local cultivars show good growth during late autumn and winter in contrast with
introductions that are winter dormant. In addition, winter active cultivars are more
productive early in spring, having a more extended period of utilisation around the year.
Thus, BFT introductions that can show an adequate degree of performance need to be
winter active. Between local cultivars, San Gabriel showed an earlier spring production
than INIA Draco.
Local cultivars are semi-erect to erect BFT types, in contrast with introduced cultivars
that are semi-prostrate types. The introduced cultivars showed a high concentration of
forage mass in low strata, with a high residual leaf area remaining for regrowth. It is
suggested that the introduced cultivars might have had a better balance between the
distribution of plant tissues above and below-ground, in contrast with local cultivars
where the high ratio of above-ground mass to below-ground mass introducing the risk
of poor persistence. This advantage on plant morphology of introduced cultivars could
not be expressed because of the winter dormancy that these materials showed in the
conditions of Uruguay.
The more intensive defoliation increased forage production during the two years, with
no significant decline in plant density. These results suggest that for the case of
prostrate cultivars, the remaining leaf area in lower strata could maintain regrowth rates
and increase the number of new sites for regrowth by more shoots. In the case of erect
types that presented a reduced leaf area independent of cutting height evaluated (4 or 8
cm), the extended periods of rest (autumn to spring) together with extended intervals
(40 days) between cutting allowed plants to rebuild levels of root reserves to an
Chapter Five 153
effective regrowth under intense defoliation. The initial time of defoliation during the
year of establishment did not affect persistence ofBFT.
In general, all cultivars suffered a strong decline in plant density during the second year,
which was attributed to drought conditions rather than disease incidence.
The nutritive value of BFT can be considered adequate in all cultivars in terms of
digestibility, protein content, fibre and condensed tannins. When BFT is considered for
stockpiling during late spring to mid summer, local cultivars are more suitable than
introduced by the high levels of forage accumulation. Because the differences in spring
growth, San Gabriel appears more adapted to accumulation in late spring and Draco for
mid summer, whilst retaining high nutritive value.
5.7 REFERENCES
ACuNA, H. ( 1 995). Comparaci6n de variedades de tres especies del genero Lotus
(Lotus corniculatus L. , Lotus uliginosus Cav. y Lotus tenuis Wald et Kit.) en
suelos de aptitud arrocera. (Comparison of three species of Lotus genus (Lotus
corniculatus L., Lotus uliginosus Cav. y Lotus tenuis Wald et Kit.) in soils
adapted for rice production). INIA Chile. Revista Agricultura Tecnica 58( 1 ) : 7-
14 . ISSN: 0365-2807.
AGRESEARCH GRASSLANDS ( 1 995). The Grasslands range of forage and
conservation plants. AgResearch Grasslands, Palmerston North. New Zealand.
ALISON, M.W. AND HOVELAND C.S. ( 1 989). Birdsfoot trefoil management. H. Yield, quality and stand evaluation. Agronomy Journal 81(5): 745-749.
AL TIER, N. ( 1 988). Enfermedades de plantas forrajeras. (Diseases in forage plants). In
lomada de Forrajeras (mimeo). Resumen de trabajos. CIAAB, Estaci6n
Experimental La Estanzuela. Colonia, Uruguay. pp. 4- 1 0.
BARRY, T.N. ( 1 989). Condensed tannins: their role in ruminant protein and
carbohydrate digestion and possible effects upon the rumen ecology. The role of
Chapter Five 154
protozoa and fungi in ruminant digestion. Eds. Nolan, IV.; Leng, R.A. and
Deneyer, DJ .. Pernanbul Books, Armidale , Australia. pp. 1 53 - 1 69.
BEUSELINCK, P.R. AND STEINER, 1.1. ( 1 995). Registration of ARS-2620 birdsfoot
trefoil . Electronic address: Ittic/tektranldatal000006/331 0000063374. Html.
BUXTON, D.R., HORNSTEIN, 1 .S . , WEDIN, W.F. AND MARTEN, G.C. ( 1 985).
Forage quality in stratified canopies of alfalfa, birdsfoot trefoil, and red clover.
Crop Science 25: 273-279.
CARAMBULA, M., BERMUDEZ, R. AND AYALA, W. ( 1 996). Caracteristicas del
Lotus pedunculatus Maku. (Characteristics of Lotus pedunculatus Maku).
Producci6n Animal. Serie Actividades de Difusi6n No. 1 1 0. INIA Treinta y
Tres. pp. 33-43 .
CARAMBULA, M., CARRIQUIRY, E. AND AYALA, W. ( 1994). Siembra de
Mejoramientos en Cobertura. (Establishment of oversown pastures). In Boletin de
Divulgaci6n No. 46. 1 994. INIA. ISBN: 9974-38-01 5-4. 20 pg. lunio 1 994.
CHARLTON, IF.L.; WILSON, E.R.L. AND ROSS, M.D.( 1 978). Plant introduction
trials. Performance of Lotus corniculatus introductions as spaced plants in
Manawatu. New Zealand Journal of Experimental Agriculture 6:20 1 -206.
DUELL, R.W. AND GAUSMAN, H.W. ( 1 957). The effect of differential cutting on the
yield, persistence, protein and mineral content of birdsfoot trefoil . Agronomy
Journal 49: 3 1 8-3 19 .
FORMOSO, F. ( 1 993). Lotus corniculatus. 1 . Performance forrajera y caracteristicas
agron6micas asociadas. (Productive performance and agronomic characteristics).
Serie Tecnica No. 37. INIA Uruguay. ISBN: 9974-556-69-4 . 20 p.
FRAME, 1 . ; CHARLTON, 1.F.L. AND LAIDLAW, A.S. ( 1 998). B irdsfoot trefoil and
greater lotus. Chapter 3 and 6. In Temperate Forage Legumes. CAB
International, Wallingford. ISBN 0-85- 1 99-2 14-5 . pp. 245-27 1 .
GARDNER, A.L.; CENTENO, G.A.; DE LUCIA, G.R. AND ALBURQUERQUE,
H.E. ( 1 968). Comportamiento de once variedades de Lotus corniculatus en La
Estanzuela. (Behaviour of eleven cultivars of Lotus corniculatus at La
Estanzuela). MGA. CIABB. Boletin Tecnico No. 8 . 23 p.
GOERING, H.K. AND VAN SOEST, P.l. ( 1 970). Forage fiber analysis. U.S.D.A. -
A.R.S. Agricultural Handbook No. 379.
Chapter Five 155
GREUB, L.J. AND WEDIN, W.F. ( 1 97 1 ). Leaf area, dry-matter accumulation, and
carbohydrate reserve level s of birds foot trefoil as influenced by cutting height.
Crop Science 1 1 : 734-738 .
LI , B . AND BEUSELINCK, P.R. ( 1 996). Plant genetic resources. Rhizomatous Lotus
corniculatus L. : Morphology and anatomy of rhizomes. Crop Science 36: 407-
4 1 1 .
NELSON, c.J. ; AND SMITH, D. ( 1 968). Growth of birdsfoot trefoil and alfalfa. n. Morphological development and dry matter distribution. Crop Science 8: 2 1 -25 .
NORBERG, S . ( 1 999). Steadfast birdsfoot trefoil . The first rhizomatous trefoil cultivar.
Missouri Agricultural Experiment Station. MU College of Agriculture, Food and
Nutrition Resources. Hundley-Whaley Farm. Field day Report Internet address:
www.aes.missouri.edulhundwhal/fielday/page I 4.stm
PIERRE, J.J. AND JACKOBS, lA. ( 1 953). The effect of cutting treatments on
birdsfoot trefoil . Agronomy Journal 45: 463-468.
REBUFFO, M. and ALTIER, N. ( 1 996). Lotus corniculatus L. LE 65-56 (INIA Draco,
a posteriori). Boletin Interno . (Lotus corniculatus L. LE 65-56. Internal report).
Programa Pasturas. INIA La Estanzuela, Uruguay. 6 p.
SAS INSTITUTE ( 1 990). SAS/STAT User 's Guide, Version 6. Cary, NC: SAS Institute. SMITH, D. AND NELSON, C.J. ( 1 967). Growth of birdsfoot trefoil and alfalfa. 1.
Responses to height and frequency of cutting. Crop Science 7: 1 30- 1 33 .
TERRILL, T.H., ROWAN, A.M., DOUGLAS, G.B. AND BARRY, T.N. ( 1 992).
Determination of extractable and bound condensed tannin concentrations in
forage plants, protein concentrate meals and cereal grains. Journal of the Science
of Food and Agriculture 58: 32 1 -329.
TILLEY, lM.A. AND TERRY, R.A. ( 1 963). A two-stage technique for the in vitro
digestion of forage crops. Journal afthe British Grasslands Society 18: 1 04- 1 1 1 .
WAGHORN, G.C., JONES, W.T., SHELTON, LD. AND McNABB, W.C. ( 1 990).
Condensed tannins and the nutritive value of herbage. Proceedings of the New
Zealand Grassland Association 51 : 1 7 1 -1 76.
WARREN WILSON, J. ( 1 963). Estimation of foliage denseness and foliage angle by
inclined point quadrat. Australian Journal of Botany 1 1 : 95-1 05 .
Chapter Five 156
WHITE, J.G.H. AND LUCAS, W.J. ( 1 990) Management of luceme in the cool season.
Proceedings of the New Zealand Grassland Association 52 : 4 1 -43.
WIDDUP, K.H. ; KEOGHAN, lM.; RYAN, D.L. AND CHAPMAN H. ( 1 987).
Breeding Lotus corniculatus for South Island tussock country. Proceedings of
the New Zealand Grassland Association 48: 1 1 9- 1 24.
Chapter Six 157
6. FORAGE PRODUCTION AND PERSISTENCE OF
BIRDSFOOT TREFOIL (Lotus corniculatus L.) IN
MIXTURE WITH WHITE CLOVER IN RESPONSE TO
DIFFERENT STRATEGIES AND INTENSITIES OF
DEFOLIATION'"
6. 1 ABSTRACT
6.2 INTRODUCTION
6.3 MATERIALS AND METHODS
6.3 . 1 Measurements
6.3 .2 Statistical analysis
6.4 RESULTS
6.4. 1 Climate conditions during experiment
6.4.2 Herbage production and quality
6.4.3 Plant density
6 .4.4 Plant morphology
6.4.5 Seed production
6.4.6 Seed soil reserves
6.4.7 Seedling emergence
6.5 DISCUSSION
6.6 CONCLUSIONS
6.7 REFERENCES
'" Part of the results in sections 6.4.5 to 6.4.7 have been accepted for publication in
Proceedings of XIX International Grassland Congress, Sao Paulo, Brazil (Ayala et al. ,
200 1 ).
Chapter Six
6.1 ABSTRACT
158
The effects of grazing management on herbage production, plant density, seed
production, seed bank size and seedling emergence of Lotus corniculatus L. San Gabriel
(BFT) and Trifolium repens cv. Zapican (WC) oversown swards were evaluated from
April 1 998 to March 2000 at INIA Treinta y Tres, Uruguay, South America. A
complete randomized block design with 4 replicate blocks was used, in which 4 grazing
strategies (grazing all year (S 1 ), summer spelling for seed production (S2), winter rest
plus summer spel ling (S3) and autumn rest plus summer spel ling (S4» , were combined
with two defoliation intensities (4 and 1 0 cm height postgrazing residuals). Plots of 1 1 0
m2 were grazed monthly by sheep.
The herbage accumulation varied from 7.7 to l O t DM/ha/year, legume contribution
being 52-54% of total. The total herbage accumulation was improved by autumn rest
(S4) during the two years and by winter rest (S3) in Year 2. Lax grazing ( 1 0 cm)
increased total accumulation 1 8-2 1 % over intensive grazing (4 cm). BFT contribution
(2.2 to 2 .3 t DM/ha/year) was affected by grazing management on Year 1 and 2,
increased if swards were lax defoliated ( 1 0 cm) or managed under extended rest period,
particularly an autumn rest (S4). WC contribution declined from 3 .0 t DM/ha in Year 1
to 1 .9 t DM/ha in Year 2, affected by extreme drought conditions. WC contribution was
increased by autumn rest (S4) and lax defoliation ( 1 0 cm).
BFT plant density was unaffected by grazing management in Year 1 , but it was reduced
by intensive defoliation (4 cm) and by grazing strategies S I and S2 between May and
December of Year 2. BFT stand was reduced 45% during the second summer and WC
growing points almost disappeared, irrespective of previous grazing management. The
seed production of BFT and WC in Year 2 was 1 3% and 2 % of production obtained in
Year 1 respectively. Summer spelling for seed production improved seed yield,
especially in BFT. Severe defoliation (4 cm) reduced seed inputs drastically (46% in
BFT and 64% in WC). BFT seed production was improved by winter rest (S3) reaching
Chapter Six 159
1 1 1 1 0 viable seeds/m2. In WC, seed production was improved by autumn rest (S4),
producing 1 1 360 viable seeds/m2. Potential seedling emergence from the soil seed
bank, between June and December, was 44% and 3 5% in BFT and WC, respectively.
Seedling emergence of BFT increased under high seed production levels (S3), and also
under intensive grazing during autumn and winter of 1 999. There were no effects on
WC seedling emergence. Emergence from the soil seed bank was 5- 1 3% and 4-7% in
BFT and WC respectively. The soil seed bank can preserve seedling recruitment rates in
the short term, but maintenance of species balance will depend on spel ling management
for seeding.
The results of this study indicate that grazing management of BFT/WC mixtures to
increase annual herbage accumulation should consider a rest during autumn and lax
grazing ( 1 0 cm). To promote seed production of legumes for future recruitment a
summer seed spelling period of 2 months is recommended in years with a decline in
stand density. Seedling recruitment is associated with soil seed reserves, but it has a low
efficiency under sward competition, demanding additional management strategies to
increase recruitment of new individuals.
Keywords: Lotus corniculatus, Trifolium repens, herbage production, persistence, seed
production, soil seed reserves, seedling emergence.
Chapter Six
6.2 INTRODUCTION
160
The oversowing of Lotus corniculatuslTrifolium repens mixtures (BFT/WC) in native
grasslands of the eastern region of Uruguay can improve forage production from 3.4 t
DMlha to 8.6 t DM/ha, increase organic matter digestibility of forage from 520 to 650
g/kg and nitrogen content from 2 1 to 29 g/kg (Ayala and Canimbula, 1 995).
Recommended management strategies to improve winter forage availability involve an
early autumn deferment for periods of 60-80 days for BFT/WC mixtures (Canimbula
and Ayala, 1 995). Rotational grazing, using rest periods around 50 days between
successive defoliations in autumn and winter, and 30 days in spring and summer
encouraged productive potential and persistence of BFT/WC mixtures. Experimental
results show liveweight gains between 390 to 550 kglha/year for five years under mixed
grazing with lambs and steers, multiplying by 6 times the levels of productivity of
rangeland (Ayala and Canimbula, 1 995).
The possibility to reduce establishment failures, complementary growth cycles of BFT
and WC, and the potential advantages of condensed tannins of BFT to the nutritive
value of mixed forages (Waghorn and Shelton, 1 997) are some of the proposed
advantages for the use of mixed legume pastures. However, these advantages tend to be
misleading when management decisions are considered. Different ferti lization
requirements (Ayala and Bermudez, 1 992; Canimbula et al., 1 994), and different
defoliation or rest period requirements (Canimbula and Ayala, 1 995), make it difficult
to develop management strategies in order to maintain an adequate balance between
BFT and WC in sward composition.
The effects of late autumn and winter defoliation could be considered as detrimental for
BFT plant survival, and the advantages of autumn deferment for subsequent forage
accumulation have already been demonstrated (Ayala and Carambula, 1 995). Effects of
disease incidence on BFT is an additional factor for stand reductions (Altier, 1 997),
Chapter Six 1 6 1
making i t necessary to encourage seed production to incorporate new individuals into
the plant popUlation.
In general, BFT is considered as short-l ived perennial legume, and strategies to improve
persistence using traditional BFT plant types are based on the recruitment of new
individuals from the soil seed bank. BFT and WC seeds are small and are produced
from late spring to summer. Roberts and Bodrell ( 1 985) determined that seedlings of
BFT and WC emerged mainly in spring, emergence patterns being reduced after 3 years
in BFT but maintaining a more regular emergence flux in WC. BFT did not show
surviving viable seeds after 5 years, contrasting with WC in which some seeds remained
viable after 5 years (Roberts and Boddrell, 1 985). Variations in patterns and magnitude
of seedling emergence between species suggest that the balance between species in the
sward can be easily altered. The final stage constitutes the seedling establishment phase,
requiring that seeds be in a proper physiological state for germination, and under
optimal environmental conditions to contribute positively to the dynamics of the plant
population of interest (Buhler et al. , 1 997).
Aspects related to dynamics of BFT and WC populations are being partially
investigated in grasslands systems of Uruguay. Recent studies on BFT popUlation
dynamics showed that, by managing recruitment processes, BFT densities could be
maintained in association with native grasses (Olmos, 1 996). On the other hand, Arana
and Pifieiro ( 1 999), working with WC, determined soil seed banks between 2600 to
1 2000 seeds/m2, seedling emergence of 200 seedlings/m2, but levels of seedling
establishment were l imited by a low seedling survival during summer. Thus, seedling
recruitment was not an effective mechanism to increase WC persistence. Also, the poor
p�rsistence of WC (3 to 4 years for conventional sowing) is associated with high death
of stolons in summer, resulting from high temperatures and water deficit (Canlmbula,
1 977 cited by Arana and Pifieiro, 1 999).
The objectives of this study were to evaluate the effect of different grazing strategies
and intensities on the production and persistence of BFT in mixed grass/legume swards
and quantify the effects of autumn or winter rest on production and plant survival of
BFT. A more detailed understanding of nutritive value of BFTIWC mixtures is
Chapter Six 162
expected to elaborate more detailed grazing plans in the conditions of Uruguay. The
potential of seed production of the species under grazing and the role of the soil seed
bank for natural reseeding were explored. Knowledge of these processes will lead to the
development of more refined management for BFTIWC mixtures in the conditions of
Uruguay.
6.3 MATERIALS AND METHODS
The study was carried out at Palo a Pique Research Unit, INIA, Treinta y Tres, Uruguay
(latitude 33° 54' S , longitude 54° 38 ' W), from April 1 998 to March 2000. The pasture
was a mixture of Lotus corniculatus cv. San Gabriel (8 kg/ha) and Trifolium repens cv.
Zapican (4.5 kg/ha) associated with native grasses (mainly perennial C4 grasses) and
established in May 1 996 by oversowing. The fertilization history was 26 kg P/ha in
1 996, 1 7 kg P/ha in 1 997 and 1 3 kg P/ha in 1 998 and 1 999 using superphospate (N-Ptotal
-P soluble-K; 0-2 1 -23-0). The soil type was a fine, thermic, mixed, vertic Argiudoll (ARS
USDA classification, Femando Garcia, personal communication) with moderate
fertility, and the soil characteristics reported in Table 6- 1 . During summer 1 998, pasture
was rested for seed production and was grazed in early March with cattle. After grazing,
the experimental area was cut uniformly at 1 0 cm height with a lawn mower.
Table 6-1 . Soil nutrients levels at experimental site in Palo a Pique Research Unit.
(Source: Soils Laboratory, INIA La Estanzuela, Uruguay).
Soil testing dates Soil depth pH (cm) (H2O)
April 1 998 0 - 7.5 5 .4 7 .5 - 1 5 5 .6
March 1 999 0 - 7.5 5 .3 7.5 - I S 5 .6
Organic carbon (%) 2.9 1 .5
2.6 1 .3
P (Bray I ) {Mg Pig soiQ
6.6 1 .7
6 . 1 2.2
K (me/1 OOg soil)
0.4 0 . 1
0.3 0 . 1
In April 1 998, four defoliation strategies and two defoliation intensities were combined
in a complete randomised block design with four replicates in plots of 1 1 0 m2•
Chapter Six 163
Defoliation strategies (Figure 6- 1 ) were grazing all year (S 1 ), summer spelling for seed
production (S2), winter rest plus summer spelling (S3) and autumn rest plus summer
spelling (S4) . Defoliation intensities were specified as post-grazing sward height of 4
or 1 0 cm.
Mar Apr May Jun Ju\ Aug Sept Oct Nov Dec Jan Feb
Autumn Winter Spring Summer
S I .
S2. ------1-----+------;. . . . . . . . . . . . . . . . . . . .
S3. ------+ . . . . . . . . . . . . . . . . . . . . . . . 1-------;. . . . . . . . . . . . . . . . . . . .
S4. . . . . . . . . . . . . . . . . . . . . . . 1-------+------+ . . . . . . . . . . . . . . . . . . .
Grazing Rest
Figure 6-1 . Description of grazing strategies applied on the birdsfoot trefoil-white
clover mixture from April 1998 to March 2000. Each strategy was defoliated at 4
and 10 cm height.
During grazing cycles, plots were grazed monthly using mature Corriedale ewes ( 1 5-
20/paddock) for short periods of time « 1 2 hours). Animals were removed overnight to
reduce faeces accumulation and rest areas. Sward height pre and post-grazing was
controlled by taking 50 readings/plot (Plate 6- 1 ) . Because sheep tend to graze hard
around fences, no samples were taken inside a boundary of one meter around plots.
Initial sward heights were established on 1 April 1 998 using a lawn mower, and plots
were mown again in March 1 999 after grazing to trim rejected mature grasses (mainly
Paspalum dilatatum and Sporobolus indicus).
When specifically referred, the seasons over the year were defined as autumn (from
March to May), winter (from June to August), spring (from September to November)
and summer (December to February).
Chapter Six 164
6.3.1 Measurements
Forage production and quality
Pre-grazing and post-grazing herbage mass were measured by cutting to ground level
with an electric shearing hand-piece two 500 x 200 mm quadrats per plot. Samples were
washed to remove residues. Those quadrat areas were marked to avoid repeat sampling.
Four values of sward height in each cut quadrat were recorded using a ruler.
The paired forage samples were individually weighed and then bulked for analysis. A
subsample was oven-dried at 60 °C for 48 hours for dry matter determination, and
another subsample was separated by hand into components (birdsfoot trefoil, white
clover, grasses and weeds), which were then oven-dried. The dead material was
included in each respective category with the green material .
Forage quality assessment ofpre-grazing herbage mass (bulk samples) included in vitro
organic matter digestibility (OMD) (Tilley and Terry, 1 963), nitrogen by Micro
Kjeldahl and acid detergent fibre (ADF) (Goering and Van Soest, 1 970).
Plant population
The plant population of BFT was recorded in May, September and December each year,
and at the end of the evaluation (March 2000). Counts were made in two fixed quadrats
per plot ( 1 x 0. 1 m each). All BFT plants with 5 or more true leaves were recorded as
adults or established plants. On two occasions (December 1 999 and March 2000)
growing points of white clover were monitored in each quadrat.
Plant morphology
In May and December of 1 998 and 1 999, five BFT plants were dug out to 250 mm
depth in each plot, and manually washed to remove soil and other components. In each
plant, the number of primary shoots and secondary shoots was recorded (see Chapter 3
for definitions). The diameter of the main root was measured at a section cut 1 0 mm
Chapter Six 165
below the l evel of insertion of primary shoots. Crown and roots (> 2 mm diameter)
were oven-dried at 60 °C for 48 hours for dry weight.
Seed production
Seed production was monitored from December 1 998 to February 1 999 (Year 1 998-
1 999) and from December 1 999 to February 2000 (Year 1 999-2000), in two fixed
quadrats per plot of 0. 1 m2 each. BFT mature pods and WC mature heads were
collected regularly over the season, dried and threshed. Measurements included number
of inflorescences, seed yield, seed number and 1 000 seed weight and germination tests
(IST A, 1 985).
Soil seed reserves
In April 1 998, March 1 999 and March 2000, six soil cores per plot (22.9 cm2 x 5 cm
depth) were taken randomly, and legume seeds recovered by a process that included
hand crumbling, sieving, air flow and addition of ethylene-cloride (C2CI4) (Prestes,
1 995 ; Appendix 2). Seeds were then hand sorted, weighed, counted and germination
tests performed (IST A, 1 985). Preliminary studies showed that in samples deeper than 5
cm the seed density was extremely low and the possibility of germination negligible.
Seedling emergence
From June to December 1 998, potential seedling emergence was checked regularly
from four soil cores collected per plot (22.9 cm2 x 5 cm depth) and placed in a complete
randomised design in an adjacent area maintained free of ground cover. Germinated
seeds were removed weekly. From March to August 1 999, seedling emergence was also
examined weekly in one fixed quadrat (0. 1 m2) per plot.
6.3.2 Statistical analysis
All data presented were evaluated using the analysis of variance of the General Linear
Model (GLM) of SAS (SAS Institute Inc., 1 990) for a balanced randomised complete
Chapter Six 166
block design. Forage production and quality data were grouped by season for analysis
and presentation. Plant population, plant morphology and seed production patterns were
analysed using a 'repeated measures analysis' .
Plate 6-1 . (a) Sheep grazing birdsfoot trefoil/white clover oversown mixture in the
experimental site at Palo a Pique Research Station, Treinta y Tres, Uruguay. (b)
Postgrazing sward heights were recorded to maintain contrasting defoliation
intensities of 4 and 10 cm.
Chapter Six
6.4 RESULTS
6.4.1 Climate conditions during experiment
167
Climate conditions during the experiment showed contrasting patterns for the two years,
particularly in rainfall occurrence, the year 1 999-2000 being extremely dry. Monthly
rainfall , soil temperatures and soil water balance of the experimental site are reported in
Chapter 5 of this thesis (see Table 5 -3 and Figure 5 - 1 ).
6.4.2 Herbage production and quality
The herbage measurements were done between May 1 998 and March 2000. Year 1
included the herbage production of 1 1 months from May 1 998 to March 1 999, and Year
2 included herbage production between March 1 999 and March 2000. Pre and post
grazing sward heights, herbage accumulation (defined as 'L(pregrazing(n+l) -
postgrazing(n))) and herbage quality parameters are reported by season.
6.4.2.1 Pregrazing and postgrazing sward height
Pregrazing sward heights varied between and across seasons, reflecting differences in
growth or accumulation periods (Table 6-2). Postgrazing sward heights showed that
contrasts between defoliation intensities were maintained over the experimental period
(Table 6-2).
Chapter Six 168
Table 6-2. Seasonal average heights (cm) of pre and post-grazing herbage (showed as pre and post) and standard deviation values in brackets for all treatment combinations.
SI- 4 CM SI- IO CM S2-4 CM S2- I O CM S3-4 CM S3- I O CM S4-4 CM S4-IO CM
Autumn Pre 1 0 (3 . 1 ) 1 4 (3.0) 9 (2.8) 13 (2.8) 1 0 (3.2) 14 (3.0) 21 (4. 1 ) 22 (4.0) 1 998 Post 4 (1 .5} 1 0 (2.0} 4 (1 .6} 1 0 (2 .0} 4 (1 .4} 1 0 (1 .9} 4 (2. 1} 9 (1 .9}
Winter Pre 9 (4.3) 13 (4.8) 9 (4. 1 ) 1 2 (4.4) 1 0 (6.0) 1 7 (7.4) 1 0 (3.9) 13 (5.0) 1 998 Post 4 (1 .3} 9 (2. 1} 4 (1 .3} 9 (1 .9} 4(1 . 1} 1 0 (1 .6} 4 (1 . 5} 9 (1 .9}
Spring Pre 1 0 (5. 1 ) 1 4 (5.2) 10 (5.7) 1 5 (6.8) 1 0 (4.8) 16 (5.7) 1 1 (4.4) 1 5 (5.5) 1 998 Post 5 (2.5} 1 0 (2.7} 5 (3.2} 1 0 (2.7} 5 (2.0) 1 0 (2 .4} 5 (2.3} 1 0 (2.9)
Summer Pre 1 1 (4 . 1 ) 1 3 (4.9) 23 (5 . 1 ) 2 3 (5.2) 2 1 (4.9) 23 (5.8) 23 (6.2) 24 (7. 1 ) 1 999 Post 5 (2.6} 1 1 (3.3} 4 (0.9} 1 0 (1 . 1) 4 (0.8} 1 0 (0.9} 4 (I .O} 1 0 (0.9}
Autumn Pre 8 (2.5) 13 (3. 1 ) 9 (2.5) 14 (2.9) 8 (2.2) 1 4 (2.9) 1 1 (2.8) IS (3.6) 1 999 Post 5 (1 .3} 1 0 (1 .8} 5 (0.9} 1 0 (4.2} 5 (I .O} 1 0 (1 .6} 4 (1 .6} 1 0 (1 .8}
Winter Pre 7 (2.3) 1 2 (4.4) 7 (2.0) 1 2 (4.4) 13 (4. 1 ) 1 9 (5.2) 8 (3.0) 1 4 (5.8) 1 999 Post 4 (1 .2} 1 0 (2.3} 4 (1 .3} 1 0 (1 .2} 5 (1 . 1} 1 0 (1 .2} 4 (1 .3} 10 (2.0}
Spring Pre 9 (2.7) I S (5.3) 8 (2.8) IS (3.8) 1 0 (2.7) 16 (6. 1 ) 1 0 (2.7) IS (3.9) 1 999 Post 4 (1 . 1} 1 0 (1 . 5} 4 (1 .4} 10 (1 .4} 4 (1 .2} 1 0 (1 .3} 4 (1 .3} 1 0 (1 . 5}
Summer Pre 7 (2. 1 ) 1 3 (2.6) 1 5 (5.6) 17 (6.0) 14 (5. 1 ) 1 5 (4.9) 13 (5. 1 ) 1 6 (6.0) 2000 Post 4 (1 .2} 1 0 (1 .7} 4 ( 1 .0) 1 0 (3.6} 4 (2.5} 1 0 (1 .2} 4 (1 .2} 1 0 (2.3}
Each height value reported includes the average of 50 records/plot in four blocks repeated three times over the season, excepting periods of rest or spelling when information is reported once at the end of the season. Autumn 1 998 included only two measurements because evaluation started in May.
6.4.2.2 Annual and seasonal herbage accumulation
The average of herbage accumulation in Year 1 reached 100 1 5 kg DM/ha, BFT and WC
contributing 22% and 30% of the total, respectively. Grasses were 44% of the total, and
Paspalum notatum, Paspalum dilatatum, Sporobolus indicus, Chloris spp. , Vu/pia
australis, Gaudinia Jragilis and Lolium multiflorum were the main contributing species.
Weeds were less than 5%, with Eringium horridum the most frequent. The general
average of seasonal accumulation was 1 8, 1 3 , 4 1 and 28% for autumn, winter, spring
and summer respectively. The autumn contribution only included two months (April
and May) because the trial started in April 1 998. There were significant differences in
total herbage accumulation in Year 1 , between strategies (P<O.O 1 ) and intensities of
defoliation (P<O.O I ), but there were no interaction effects between main factors (Table
6-3). The S4 strategy accumulated more forage than the others, and S I was the least
productive. The lax grazed plots ( l O cm) produced 1 8% more forage than the
intensively grazed plots (4 cm).
Chapter Six 169
Table 6-3. Annual and seasonal herbage accumulation (kg DM/ha) of an oversown
birdsfoot trefoil/white clover mixture managed under different strategies and
intensities of grazing during the third and fourth year after establishment.
Year 1 Year 2 Autumn Winter Spring Summer Total Autumn Winter Spring Summer Total
S I - 4 cm 1 575 920 2505 2495 7505 1 635 1 040 2270 620 5570 S I - I O em 1 555 1 225 2655 3585 9025 1 625 1 320 3 1 05 875 6920 S2 - 4 cm 1 625 1 285 3075 2380 8360 2020 1 35 5 2990 635 7005 S2 - 1 0 em 1 960 1 570 3475 3025 1 0025 2475 1 930 3670 565 8645 S3 - 4 cm 1 585 1 05 5 4465 2330 9430 2025 1 475 3245 545 7285 S3 1 0 cm 1 985 1 080 6095 3 1 75 1 2345 2490 1 75 5 4480 535 9265 S4 - 4 cm 1 780 1 565 4545 2825 1 07 1 5 2 1 60 1 800 3245 505 77 10 S4 - I O cm 2 1 5 5 1 805 6240 2 5 1 2 1 2 7 1 5 2 1 70 1 765 4720 575 9230
SEM (S x J) 1 02 1 02 1 63 1 98 270 1 55 8 1 1 03 80 1 95 Signif. (S x I) NS NS • • • • NS NS • • • • NS NS
Signif. (S) NS •• • • NS • • • • • • * * NS • •
Silionif. (Q * . NS * * * * • • N S • • • N S * .
**, P<O.O l ; NS, not significant, SEM (S x I), standard error of the mean of the interaction strategy x
intensity of defoliation; n=4; S ignif., significance; S, strategy of defoliation; I , intensity of defoliation
Autumn herbage accumulation in Year 1 was affected only by grazing intensity
(P<O.O 1 , Table 6-3), production being 1 4% higher in plots grazed at 1 0 cm height. In
contrast, herbage accumulation during winter of Year 1 was affected only by grazing
strategy, the autumn rest strategy (S4) being the most productive. A significant
interaction strategy x intensity of defoliation was observed (P<O.O 1 ) in spring of Year 1
(Table 6-3). Strategies S3 and S4 showed an increase in accumulation when intensity of
grazing changed from 4 to 10 cm but accumulation remained unchanged in S I and S2.
In summer of Year 1 , there was an interaction of strategy x intensity (P<O.O 1 , Table 6-
3), strategies increasing accumulation if lax defoliated with the exception of S4 that was
not affected by defoliation intensity.
In Year 2, the average of herbage accumulation was 7705 kg DMlha. The yield was
composed of 25 , 29, 40 and 6% of BFT, WC, grasses and weeds respectively. There
was a decline in WC and grasses production that affected total accumulation. In
contrast, BFT accumulation was similar to that in Year 1 , reflecting a high tolerance of
BFT to drought. Herbage accumulation in Year 2 was affected by strategies (P<O.O l )
and intensities o f defoliation (P<O.O l ), as occurred in Year 1 (Table 6-3). The general
average of seasonal distribution was 27, 20, 45 and 7% for autumn, winter, spring and
summer respectively. There were no interaction effects for total production (Table 6-3).
Herbage accumulation was improved by rest period independently of the timing,
Chapter Six 170
ranking being S3 and S4> S2>S l . The lax grazing ( 1 0 cm) remained more productive
(2 1 % higher) than the intensive grazing (4 cm) as was observed in Year 1 .
The herbage accumulation in autumn of Year 2 was significantly affected (P<O.O l ) by
grazing strategies, the more intensive system (S 1 ) reduced accumulation by 27% when
compared with strategies that included any kind of rest during the year (S2, S3 or S4).
Winter accumulation was affected significantly by the interaction strategy x intensity
(P<O.O l , Table 6-3), and herbage accumulation increased under lax defoliation for S I ,
S2 and S3 strategies, but S4 was not affected by defoliation intensity. In spring, a
significant interaction was observed for strategy x intensity of defoliation (P<O.O 1 ,
Table 6-3) . Plots grazed laxly ( 1 0 cm) had an accumulation 26% higher than those
defoliated more intensively, and strategies that had more extended rest were more
productive (S3 and S4>S2>S 1 ). There were no significant treatment effects on summer
production (Table 6-3).
6.4.2.3 Species contribution
The seasonal and annual accumulation of BFT, WC and grasses during the two years is
reported in Tables 6-4 to 6-6. Weeds made a minor contribution to total herbage
accumulation between 5 -6% during the two years and data are not presented.
Birdsfoot trefoil
BFT accumulation was similar in the two years (2 165 and 2270 kg DM/ha for Year 1
and Year 2 respectively). Consistently during the two years, there was a significant
interaction strategy x intensity of defoliation on total BFT accumulation (P<O.O I , Table
6-4). In Year 1 , herbage accumulation increased if BFT grazed at 1 0 cm height instead
of 4 cm in all systems excepting S3. In Year 2, herbage accumulation was improved
30% if swards were grazed at 1 0 cm rather than at 4 cm height, and also by those
grazing strategies that had more extended rest during the year. Autumn rest was more
productive than winter rest (S4>S3>S2>S 1 ).
Chapter Six 171
In Year 1 , there was a significant interaction strategy x intensity of defoliation in all
seasons (P<O.O I m all cases, Table 6-4). In autumn, there was no improved
accumulation by lax defoliation in S I , in contrast with other strategies. The same
pattern was registered in winter for S3, in spring for S2 and in summer for S4 (Table 6-
4). Summer contribution was improved in S I over the other strategies, because the
effect of seeding rest in the other strategies determined losses of BFT herbage.
Table 6-4. Annual and seasonal herbage accumulation of birdsfoot trefoil (kg
DM/ha) in a birds foot trefoil/white clover mixture managed under different
strategies and intensities of grazing during the third and fourth year after
establishment.
Year 1 Year 2 Autumn Winter Spring Summer Total Autumn Winter Spring Summer Total
S I - 4 cm 375 90 380 1 035 1 885 375 85 265 1 35 860 S I - I O em 240 3 1 0 560 1 695 2805 495 285 755 285 1 820 S2 - 4 em 4 1 5 1 60 440 440 1450 565 275 480 1 55 1475 S2 - 10 cm 600 455 590 605 2245 1 000 485 1 1 1 5 205 28 1 0 S3 - 4 cm 570 300 745 895 2505 925 665 785 1 20 2490 S3 - 1 0 cm 570 145 695 975 2385 1 205 350 1 095 1 60 28 1 0 S 4 - 4 cm 480 300 470 570 1 820 940 590 905 220 2660 S4 - I O em 575 420 660 565 2220 1 1 1 0 440 1 535 1 55 3240
SEM (S x J ) 32 20 27 59 63 88 22 44 29 96 Signif. (S x I) * * * * * * * * * * NS * * * * * * * *
Signif. (S) NS NS NS ** NS ** * * * * N S * *
Sil;init: !Q NS NS NS NS NS * * NS * * NS * *
**, P<O.Ol ; NS, not significant; SEM (S x I), standard error of the mean of the interaction strategy x
intensity of defoliation; n=4; Signif., significance; S, strategy of defoliation; I, intensity of defoliation
During autumn of Year 2, mam effects were significant (P<O.O l , in both cases),
accumulation improved when swards were defoliated laxly or received an extended rest
in the previous year. During winter, the interaction strategy x intensity was significant
(P<O.O 1 ), with strategies increasing accumulation if lax defoliated, with the exception of
S4 (Table 6-4).
In spring the interaction strategy x intensity of defoliation was significant (P<O.O l ,
Table 6-4), and main effects were also significant (P<O.O I ) . Accumulation increased
46% if lax defoliated ( 1 0 cm) over the intensive defoliation (4 cm), and in less intensive
grazed strategies (S4>S3>S2>S 1 ). In summer, there was a substantial reduction in
accumulation by drought. The interaction strategy x intensity was significant (P<O.O l ),
Chapter Six 1 72
the S I treatment increasing accumulation if lax grazed, but the other strategies did not
differ if defoliated at 4 or 1 0 cm height.
White clover
The total accumulation of WC was higher in Year 1 than in Year 2 (2970 and 1 925 kg
DMlha for Year I and Year 2 respectively). There was a significant interaction strategy
x intensity of defoliation in the two years (P<O.O 1 , Table 6-5). General patterns over
time showed that the intensity was a significant variable from the first spring to the end
of the experiment, and consistently there were differences between systems in spring
and summer of each year. In Year 1 , strategies increased total accumulation if
defoliated at 10 cm rather than at 4 cm, but S2 accumulation was not affected by
defoliation intensity. Effects of main factors were also significant, showing an increase
of 3 1 % in WC accumulation if grazed at 1 0 cm height. The autumn rest improved WC
contribution over the other treatments, the contribution being reduced by the more
extended grazing over the year.
Table 6-5. Annual and seasonal herbage accumulation of white clover (kg DMlha)
in a birdsfoot trefoil/white clover mixture managed under different strategies and
intensities of grazing during the third and fourth year after establishment.
Year 1 Year 2 Autumn Winter Spring Summer Total Autumn Winter Spring Summer Total
S I - 4 cm 275 350 680 3 1 5 1 620 3 1 5 4 1 5 6 1 5 0 1 345 SI - I O em 380 450 825 460 2 1 1 0 395 505 725 30 1 650 S2 - 4 cm 440 645 1 000 1 40 2220 420 440 705 0 1 565 S2 - 10 cm 440 590 985 245 2260 485 860 1 1 1 0 0 2460 S3 - 4 cm 420 550 1 525 1 35 2630 1 90 320 780 0 1 290 S3 - 1 0 cm 550 5 1 5 3280 280 4630 575 705 1 500 0 2785 S4 - 4 cm 660 530 2045 1 25 3400 380 545 625 0 1 545 S4 - 10 cm 800 790 3030 295 4920 530 760 1 470 0 2760
SEM (S x I) 64 72 70 26 1 2 1 28 38 45 2 58 Signif. (S x I) NS NS * * NS * * . . * * * * * * * *
Signif. (S) ** NS ** * * . . NS NS * * * * *
Si!;linif. �Q NS NS * * . . * * • • • • . . . * * *
** , P<O.O I ; NS, not significant; SEM (S x I), standard error of the mean of the interaction strategy x intensity of defoliation; n=4; Signif. , significance; S, strategy of defoliation; I, intensity of defoliation
Chapter Six 173
In autumn of Year 1 , there was an effect of grazing strategies on WC accumulation
(P<O.O I ), the autumn rest increased WC contribution over the other strategies. In winter
there were no treatment effects on WC accumulation. During spring, the interaction
strategy x intensity was significant (P<O.O 1 , Table 6-5). The accumulation of WC in S I
and S2 strategies was not affected by grazing intensity, but in S3 and S4 increased if
plots were grazed at 10 cm height. In summer of Year 1, there were significant effects
of the strategy and intensity of defoliation (P<O.O I in both cases). Accumulation of WC
was higher in the treatment S I compared with those that received a rest for spelling, and
WC increased when plots were defoliated at 10 cm height.
In Year 2, the interaction strategy x intensity was significant in autumn (P<O.O l , Table
6-6) and the effect of intensity was also significant (P<O.O 1 ). Strategies S I and S2 did
not differ in accumulation by effect of defoliation intensity, but S3 and S4 increased
accumulation if defoliated at 1 0 cm rather than at 4 cm height. During winter, the
interaction strategy x intensity was significant (P<O.O I , Table 6-6). Strategies S2, S3
and S4 increased accumulation when grazed at 10 cm rather than 4 cm, but S I was not
affected by changes in defoliation intensity. The tendency described in winter was
observed in spring (Table 6-5). In summer of Year 2, there was a decline in WC
contribution, only a minimum presence of WC was observed in S 1 - 1 0 cm during early
sampling in summer (Table 6-5).
Grasses
Grasses were less affected than legumes by grazing management. Total herbage
accumulation from grasses in Year 1 was only affected by grazing strategies (SEM 1 33 ,
P<O.05), ranking being S4>S2, S3>S 1 . In Year 2, there were no significant effects on
total production (Table 6-6).
Chapter Six 1 74
Table 6-6. Annual and seasonal herbage accumulation of grasses (kg DM/ha) in a
birdsfoot trefoil/white clover mixture managed under different strategies and
intensities of grazing during the third and fourth year after establishment.
Year 1 Year 2 Autumn Winter Spring Summer Total Autumn Winter Spring Summer Total
S I - 4 cm 895 460 1 335 1 0 1 5 3705 855 500 1 305 440 3 1 00 S I - 1 0 em 905 430 1 205 1 295 3840 635 465 1 560 505 3 1 70 S2 - 4 cm 705 435 1 5 1 0 1 645 4295 8 1 5 485 1 505 405 32 1 5 S2 - 1 0 cm 845 395 1 770 2020 5035 745 370 1 365 280 2760 S3 - 4 cm 530 1 80 2030 1 200 3940 620 435 1 490 380 2920 S3 - 1 0 cm 755 385 1 970 1 8 1 5 4925 5 1 5 665 1 695 345 3220 S4 - 4 cm 545 650 1 820 2000 50 1 5 795 485 1 635 250 3 1 65 S4 - 10 cm 685 480 2465 1 590 5225 470 4 1 5 1 565 395 2845
SEM (S x I ) 72 68 1 09 1 46 1 88 1 06 5 8 1 03 7 1 1 45 Signif. (S x I) NS NS .* NS NS NS NS NS
Signif. (S) NS NS • NS NS NS NS NS NS Sil;inif. \Q NS NS NS NS NS * NS NS NS NS * *, P<O.O l ; * , P<O.05; NS, not significant; SEM (S x I), standard error of the mean of the interaction strategy x intensity of defoliation; n=4; Signif., significance; S, strategy of defoliation; I, intensity of defoliation
In spring and summer of Year 1 there was a significant interaction strategy x intensity
of grazing (Table 6-6). In spring, only S4 increased accumulation if defoliated at 1 0 cm
rather than at 4 cm height (Table 6-6). During summer, accumulation of S4 decreased if
lax defoliated ( 1 0 cm), in contrast with other strategies that increased accumulation
under more lax defoliation. However, the results need to be carefully interpreted,
because the accumulation included three months of rest, with the implications of
maturity, and death of herbage.
6.4.2.4 Herbage quality
The in vitro organic matter digestibility (OMD), nitrogen content and acid detergent
fibre (ADF) content were analysed by season in pre-grazing herbage samples. During
Year 1 , there were significant differences in OMD (SEM 5 .5, 2.8, 2 .5 and 7.3 for
autumn, winter, spring and summer respectively, P<O.Ol in all seasons) between
grazing strategies. The highest OMD levels in autumn were observed for S4, in winter
for S3 , in spring for S4 and in summer for S I . The intensity of defoliation affected
OMD during autumn (SEM 3 .9, P<O.O l ), winter (SEM 6.2, P<O.05) and spring (SEM
1 .8 , P<O.O l ), but not in summer. The OMD when significant, was always higher in 4
Chapter Six 175
cm than 1 0 cm height. There were no interaction effects strategy x intensity of
defoliation for OMD in Year 1 (Table 6-7).
In Year 2 , a significant interaction strategy x intensity of defoliation was observed for
OMD in winter and spring (P<O.Ol in both cases). In the two seasons OMD increased if
plots were grazed intensively, with the exception of S4 in winter and S3 in spring,
where OMD was unaffected by changes in defoliation intensity. In autumn, there was a
strategy of defoliation effect (P<O.O 1 ), S4 having higher OMD than the other strategies,
as was observed in the previous year. During summer there were no significant effects
(Table 6-7).
Table 6-7. Seasonal averages of in vitro organic matter digestibility (g/kg DM) of a
birdsfoot trefoil/white clover oversown mixture, during two years.
Year 1 Year 2
Autumn Winter Spring Summer Autumn Winter Spring Summer
S I - 4 cm 580 679 534 471 566 666 5 1 0 497 S I - I O cm 552 666 5 1 9 443 56 1 633 472 496 S2 - 4 cm 5 74 673 554 3 72 559 661 528 509 S2 - 10 cm 554 672 524 336 556 6 1 3 507 5 1 2 S 3 - 4 cm 5 76 700 559 344 550 638 502 504 S3 - 10 cm 574 693 532 353 556 656 5 1 1 496 S4 - 4 cm 634 677 576 370 6 1 7 626 509 506 S4 - I O cm 59 1 661 54 1 376 607 635 5 3 1 503
SEM (S x 1 ) 7.8 5.0 3.6 1 0.4 6.3 4.0 6 . 1 8.0 Signif. (S x J) NS NS NS NS NS * * * * N S
Signif. (S) ** ** • • * * * * * * * * NS Sijjnif. �I� . * * * " NS NS * * NS NS
* *, P<O.O l ; *, P<O.05; NS, not significant; SEM (S x I), standard error of the mean of the interaction strategy x intensity of defoliation; n=4; Signif, significance; S, strategy of defoliation; I, intensity of defoliation
The nitrogen content in Year 1 was affected by grazing strategies in spring (SEM 0.4,
P<0.05) and in summer (SEM 0.3 , P<O.O l ). In spring, the nitrogen content in S4 was
higher than in S3 and S I . In summer, treatment S I showed the highest nitrogen content.
In autumn, there was a significant effect of defoliation intensity (SEM 0.3, P<0.05),
plots defoliated at 4 cm had higher nitrogen content than those defoliated at 10 cm
height. In winter, there were no differences in nitrogen content between treatments
(Table 6-8).
Chapter Six 1 76
Table 6-8. Seasonal averages of nitrogen content (g/kg DM) of a birdsfoot
trefoil/white clover oversown mixture, during two years.
Year I Year 2
Autumn Winter Spring Summer Autumn Winter Spring Summer
S I - 4 cm 30 35 27 22 2 1 25 1 4 1 4 S I - I O em 30 35 28 2 1 22 25 1 4 1 4 S 2 - 4 cm 3 1 3 5 29 16 22 26 1 4 1 5 S 2 - 1 0 cm 30 29 28 1 5 22 24 1 4 1 4 S3 - 4 cm 30 32 28 14 22 23 1 6 1 4 S 3 - 1 0 cm 30 33 28 1 5 22 24 1 5 1 4 S4 - 4 cm 32 36 30 1 3 27 27 1 5 1 4 S4 - I O em 29 35 29 14 27 28 14 14
SEM (S x l) 0.7 1 .9 0.6 0.4 0.6 0.6 0.5 0.6 Signif. (S x J) NS NS NS NS NS NS NS NS
Signif. (S) NS NS ** ** ** NS NS Sil:\nif. �Jl * NS NS NS NS NS NS NS
** , P<O.O I ; *, P<O.05; NS, not significant; SEM (S x I), standard error of the mean of the interaction strategy x intensity of defoliation; n=4; Signif., significance; S, strategy of defoliation; I, intensity of defoliation
In Year 2, nitrogen content was affected by strategies of defoliation during autumn
(SEM 4.4, P<O.O l ) and winter (SEM 2.9, P<O.O l ). In autumn, the S4 treatment had the
highest Nitrogen content. In winter, S3 and S I had higher nitrogen content than S2 and
S4.
During Year 1 , the ADF content did not differ between treatments in autumn, but it was
affected by defoliation intensity in winter (P<O.O 1 , SEM 2 . 1 ), the higher fibre content
being observed in plots lax defoliated ( 1 0 cm). In spring, there was a significant
interaction strategy x intensity of defoliation (P<0.05, Table 6-9), the S3 strategy
increased ADF if defoliated at 1 0 cm rather than 4 cm but the other strategies remained
unchanged. In summer, there was a significant effect of defoliation strategy (SEM 7.3,
P<O.O I ), S I having the lowest ADF content.
In autumn of Year 2, there was a significant effect of defoliation strategy (SEM 5 .2,
P<O.O I ), the ranking being S4< S I , S2 and S3. In winter there was a significant
interaction strategy x intensity of defoliation (P<O.O 1 , Table 6-9). Lax defoliation
increased ADF in strategies S2 and S3, but decreased in S4. In spring, the ADF content
was affected by defoliation intensity (SEM 2.9, P<0.05), increasing in swards lax
defoliated. Finally in summer of Year 2, ADF was significantly affected by defoliation
Chapter Six 177
intensity (SEM 3 .2, P<0.05), ADF fraction increasing in lax defoliated swards ( 1 0 cm)
as reported in spring.
Table 6-9. Seasonal averages of acid detergent fibre (g/kg DM) of a birdsfoot
trefoil/white clover oversown mixture, during two years.
Year I Year 2 Autumn Winter Spring Summer Autumn Winter Spring Summer
S I - 4 cm 352 33 1 340 438 385 33 1 399 4 1 1 S I - IO cm 364 34 1 343 465 395 339 426 423 S2 - 4 cm 350 331 343 549 393 308 406 417 S2 - 10 cm 366 339 356 535 388 362 422 434 S3 - 4 cm 352 328 33 1 52 1 384 372 392 4 1 1 S3 - 1 0 cm 347 340 356 5 1 7. 386 355 420 423 S4 - 4 cm 347 322 34 1 500 325 327 397 4 1 5 S4 - 1 0 cm 363 344 337 498 325 346 4 1 6 420
SEM (S x I) 7.7 4. 1 5.0 8.9 7.3 4.7 5.8 6.4 Signif. (S x I) NS NS • NS NS .. NS NS
Signif. (S) NS NS NS .. •• . . NS NS Si�nif. (I� NS . . NS NS .. .. •
** , P<O.O I ; *, P<O.05; NS, not significant; SEM (S x I), standard error of the mean of the interaction strategy x intensity of defoliation; n=4; Signif., significance; S, strategy of defoliation; I, intensity of defoliation
6.4.3 Plant density
Initially in May 1 998, plant density of BFT was uniform with an average of 86
plants/m2. Over time, a significant interaction time x defoliation strategy (P<0.05) was
observed (Figure 6-2 a), but there were no other interaction effects.
Plant density of BFT increased on average 34% from May 1998 to May 1999, but
decreased 54% from May 1 999 to March 2000. During the first year, differences were
registered in May 1999. Plant density in strategies S3 and S4 was 20% higher on
average than in strategies S I and S2 (Figure 6-2 a), and 1 0% higher in swards grazed at
1 0 cm rather than 4 cm (Figure 6-2 b).
Chapter Six
1 50
1 25
lOO
Ne u;
75 ] 0..
50
25
150
120
90 Ne
� 60
30
(a) Defoliation strategies
NS NS
___ S I
-o- S2
---'- S3
--l!r- S4
May 1 998 SeP' 1 998 Dee 1 998 May 1 999 Sep' 1 999
NS
May 1 998
NS
___ 4 cm -o- lO cm
Sep' 1 998
NS
Dee 1998
Dates
(b) Defoliation Intensities
I
I
May 1 999 Sep' 1 999
Dates
NS
Dee t 999 March 2000
NS
Dce 1 999 March 2000
1 78
Figure 6-2. Changes in BFT density (adult plants) under (a) four defoliation
strategies and (b) two defoliation intensities from April 1 998-March 2000. Vertical
bars indicate SEM (nstrategies=16, nintensities=32), and NS not significant differences at
corresponding sampling dates.
Plant density was significantly affected (in May P<O.O 1 , in September P<O.O 1 and in
December P<0.05) by defoliation strategies, S3 and S4 having higher density than S I
and S2. (Figure 6-2 a); and by defoliation intensity (in May P<O.O l , September P<0.05
and December 1 999 P<O.O l respectively) (Figure 6-2 b), the 1 0 cm height having higher
Chapter Six 179
plant density than 4 cm. The rate of plant loss was high during spring 1999 and summer
2000, coinciding with drought conditions. Population declined 45% in summer 2000,
leaving a final density of 42 plants/m2• There were no differences in final density between
treatments (Figure 6-2), resulting in an open and heterogeneous sward.
During the drought period growing points of white clover were monitored. In December
1 999, there were significant differences between defoliation intensities (29 and 1 1
growing points/m2 for 1 0 and 4 cm height respectively, SEM 6.7, P<0.05). Surviving
growing points were located mainly in areas covered by native grasses or dead material.
In March 2000, there were no growing points of WC in any of the treatments.
6.4.4 Plant morphology
Information about the number of primary shoots, secondary shoots, crown mass, root
mass and root diameter in BFT plants was analysed from May 1 998 to December 1 999,
and is presented in the sections 6.4.4. 1 to 6.4.4 .5 .
6.4.4. 1 Primary shoots
Over time, there was a decline in the general average from 6 to 4 primary shoots per
plant. There was a significant interaction time x defoliation strategy (P<O.O I , Table 6-
1 0), but not for defoliation intensity or strategy x intensity over time. There were
differences between defoliation strategies in May and December 1 998, but not during
1 999. From May 1 998, the autumn rest (S4) increased the number of shoot s/pl ant over
the other treatments (Table 6- 1 0). In December 1 998, defoliation strategies that had a
previous rest during the year (S3 and S4) had higher shoot density than systems without
rest (S 1 and S2) (Table 6- 1 0).
Chapter Six 180
Table 6-10. Number of primary shoots/plant of BFT under different strategies and
intensities of defoliation from April 1998 to December 1 999.
May 1998 December 1998 May 1999 December 1 999 SEM /Significance
Defoliation strategies
S I 5 4 3 4 S2 6 4 4 4 SEM (lime x stralegy)
S3 5 5 4 4 0.9 * * S4 8 6 4 4
SEM (n) 0.6 (8) 0.3 (8) 0.5 (8) 0.5 (8) Significance * * * * NS NS
Defoliation intensities
4 cm 6 5 3 4 SEM (time)
I O cm 6 5 4 4 0.5 NS
SEM (n) 0.4 ( 1 6) 0.2 ( 1 6) 0.3 ( 1 6) 0.4 ( 1 6) Significance NS NS NS NS
** , P<O.O I ; NS, not significant; SEM, standard error of the mean; (n) number of observations for each treatment mean
6.4.4.2 Secondary shoots
A significant time x strategy x intensity of defoliation interaction (P<O.O l ) was
observed for the number of secondary shoots (Table 6- 1 1 ), the average of shoots per
plant decreasing from 1 3 to 8 during the experimental period. Within sampling dates,
only during the first sampling was the interaction strategy x intensity significant
(P<O.05, Table 6- 1 1 ), secondary shoots increasing if plants received a rest (S4); the
other treatments did not differ if defoliated at 4 or 1 0 cm height. In December 1 998, the
height of defoliation affected secondary shoots per plant (SEM 1 .0, P<0.05), plant lax
defoliated ( 10 cm) having more secondary shoots than those hard defoliated (4 cm). The
final sampling showed significant differences by defoliation intensity (SEM 0.6, P<O.O I ),
plants lax defoliated maintaining a high density of secondary shoots.
Chapter Six 181
Table 6-1 1 . Evolution of secondary shoots (no.lplant) of BFT plants under a
combination of four defoliation strategies and two defoliation intensities from
April 1998 to December 1999.
May 1 998 December 1998 May 1 999 December 1999 SEM ISignificance
Defoliation Strategy x intensity
S l -4 cm l 3 6 6 7 SEM (lfmt 1 lInltgy 1 jfl'tMi(YJ
S I - 1 0 cm 1 1 1 1 6 l O l A * * S2-4 cm 1 2 9 7 6 S2- 1 0 cm 1 4 1 4 8 9 S3-4 cm l 3 1 2 6 6 S3- 1 0 cm 1 2 1 1 6 1 1 S4-4 cm 1 2 9 5 7 S4- 1 0 cm 23 20 1 0 9
SEM (n) 2 . l (4) 2.0 (4) l A (4) 1 .3 (4) Significance * NS NS NS
** , P<O.O l ; * , P<0.05; NS, not significant; SEM, standard error of the mean; (n), number of observations for each treatment mean
6.4.4.3 Crown mass
There was a significant interaction time x defoliation intensity effect (P<O.05), but no
effect of defoliation strategy or strategy x intensity over time (Table 6- 12). A clear
tendency to reduce crown mass of plants grazed at 4 cm height in contrast with those
grazed at 1 0 cm was observed during the last three sampling dates (Table 6- 1 2).
Differences in crown mass between defoliation strategies were found during the three
first dates (Table 6- 1 2), S4 having bigger crowns than the others. The average for all
sampling dates showed that plants of S4 strategy had crowns 53% bigger than the
average of the other strategies.
Chapter Six 1 82
Table 6-12 . Evolution of crown mass (g/plant) of BFT plants under four defoliation
strategies and two defoliation intensities from April 1998 to December 1999.
May 1998 December 1998 May 1 999 December 1999 SEM ISignificance
Defoliation strategies S I 0.8 0.5 0 .8 0.8 SEM (time x sttategy) S2 1 .0 0 .7 0.9 0.9 0. 1 2 NS S3 0.9 0.9 1 .0 1 .0 S4 1 .3 1 .2 1 .4 1 .2
SEM (n) 0 . 1 3 (8) 0.08 (8) 0. 1 0 (8) 0. 1 6 (8) Significance * * * ** NS
Defoliation intensities SEM (time x intensity) 4 cm 1 .0 0.6 0.9 0.7 0.08 * 1 0 cm 1 . 1 1 .0 1 .2 1 .3
SEM (n) 0.09 ( 1 6) 0.05 ( 1 6) 0.07 ( 1 6) 0. 1 1 ( 1 6) Significance NS * * * * **
* * , P<O.O I ; * , P<0.05; NS, not significant; SEM, standard error of the mean; Cn), number of observations for each treatment mean
6.4.4.4 Root mass
Significant interactions time x defoliation strategy (P<O.05) and time x defoliation
intensity (P<O.05) were observed for root mass (Table 6- 1 3), but there was no
interaction between main effects over time. Defoliation strategies affected root mass
significantly during the first three sampling dates (Table 6- 1 3). The more intensive
strategies (S 1 and S2) had less root mass per plant compared with more lax strategies,
especially S4. Plants grazed at 1 0 cm had consistently greater root mass than those
grazed at 4 cm height (Table 6- 1 3 ), and the differences were significant for last three
sets of dates studied.
Chapter Six 183
Table 6-13. Evolution of root mass (g/plant) of BFT plants under four defoliation
strategies and two defoliation intensities from April 1998 to December 1999.
May 1998 December 1998 May 1 999 December 1 999 SEM ISignificance
Defoliation strategies S I 1 . 1 0.9 1 . 1 0.8 SEM (time x strategy) S2 1 .2 1 .2 1 .3 0.9 0. l 3 * S3 1 .2 l A 1 .6 0.9 S4 2.2 1 .8 2.0 1 .0
SEM (n) 0.2 1 (8) 0. 1 4 (8) 0.5 (8) 0 . 1 3 (8) Significance * * * * * * N S
Defoliation intensities SEM (time x intensity) 4 cm 1 .2 0.9 1 . 1 0.7 0.09 * 1 0 cm 1 .6 1 .7 1 .9 1 . 1
SEM (n) 0. 1 5 ( 1 6) 0.09 ( 1 6) 0 . 1 2 ( 1 6) 0.09 ( 1 6) Significance NS * * * * * *
** , P<O.O 1 ; NS, not significant; SEM, standard error o f the mean; (n) number of observations for each treatment mean
6.4.4.5 Root diameter
A significant interaction time x defoliation intensity (P<O.05) was observed for root
diameter, values for BFT plants defoliated at 1 0 cm height being greater than those for
plants defoliated at 4 cm (Table 6- 14). A fast decline in root diameter of plants
defoliated at 4 cm occurred between first and second sampling, differences maintained
during the rest of the period excepting the final evaluation (Table 6- 1 4).
There were differences among defoliation strategies in December 1 998 (P<O.O 1 , Table
6- 1 4). Root diameter of BFT plants in strategies that received a previous rest (S3 and
S4) was higher than in strategies without any rest to this time (S 1 and S2).
Chapter Six 184
Table 6-14. Evolution of root diameter (mm) of BFT plants under four defoliation
strategies and two defoliation intensities from April 1998 to December 1999.
May 1998 December 1998 May 1999 December 1999 SEM ISignificance
Defoliation strategies S I 1 0 8 6 7 SEM (lime x ""'e&Y) S2 I I 8 7 8 2.8 NS S3 1 0 1 0 8 7 S4 1 2 1 0 8 7
SEM (n) 0.6 (8) 0.3 (8) 0.5 (8) 0.5 (8) Significance NS ** NS NS
Defoliation intensities 4 cm I I 8 6 7 1 0 cm I I 1 0 8 8
SEM (lime x inlen.i,>,)
SEM (n) 0.4 ( 1 6) 0.2 ( 1 6) 0.3 ( 1 6) 0.4 ( 1 6) 0.5 *
Significance NS * * * NS
**, P<O.O I ; * , P<0.05 ; NS, not significant; SEM, standard error of the mean; (n), number of observations for each treatment mean
6.4.5 Seed production
Seed production differed significantly between years. Seed yield of BFT in Year 2 was
only 1 3% of the yield in Year 1 (SEM 0.679, P<O.O I , Table 6- 1 5). Seed production of
WC was also seriously reduced during the second year, yield being only 2% of that
achieved in the previous year (SEM 0.334, P<O.O I , Table 6- 1 5). There were no
interaction effects, Table 6- 1 5 shows main effects in each species (strategy and intensity
of defoliation); analysis was done by year independently due to disparities in seed yield.
BFT seed production was significantly affected (P<O.O I ) by grazing strategies, yield for
the unspelled treatment (S 1 ) being only 6% of the summer spell treatment (S2). Winter
rest improved BFT seed production over the other rest treatments (Table 6- 1 5).
Management effects on WC were not significant, though again seed production was
lower in S I than in the other treatments. Defol iation intensity did not affect seed
production in BFT, but intense defoliation (4 cm) reduced seed production in WC to
32% of yield in swards defoliated at 1 0 cm (Table 6- 1 5).
Chapter Six 185
Table 6-15. Annual seed production (g/m2) of birdsfoot trefoil (BFT) and white
clover (WC) in mixture under different strategies and intensities of defoliation
during two years.
Year 1 Year 2 BFT WC BFT WC
Defoliation strategy S I 0.6 2.5 0.2 0.004 S2 9.7 4.3 0.9 0.024 S3 1 5.0 4.2 1 .0 0.000 S4 9.5 6.4 2.4 0.024
SEM (n) 1 .26 ( 1 6) 0.73 ( 1 6) 0.39 ( 1 6) 0.0070 ( 1 6) Significance ** NS NS NS
Defoliation intensity 4 cm 6.9 2 .0 1 .0 0.006 1 0 cm 1 0.5 6.6 1 .2 0.020
SEM (n) 0.89 (32) 0.52 (32) 0.28 (32) 0.0049 (32) Significance NS ** NS NS
General mean 8.7 4.3 1 . 1 0.0 1
** , P<0.0 1 ; NS, not significant; SEM, standard error of the mean; (n), number of observations for each treatment mean
During the second year, the seed production of BFT and WC was not affected by either
grazing strategy or the intensity of defoliation (Table 6- 1 5) . In general , the SEM values
reported in Table 6- 1 5 are high compared with treatment means during the two years, a
consequence of the heterogeneity in species distribution in a three year old oversown
pasture.
6.4.5.1 Seed yield components
The number of inflorescences/m2 in BFT was affected (P<O.O l ) by defoliation
strategies in Year 1 , the unspel led treatment (S 1 ) having only 9% of the inflorescences
of the spelled treatment (S2), and the winter rest further increased the number of the
inflorescences of BFT. In Year 2 , there were no effects of defoliation strategy on BFT
inflorescences. Defoliation intensity did not affect the inflorescences/m2 in either Year 1
or Year 2 (Table 6- 1 6). In WC, the number of inflorescences was affected (P<O.05) by
Chapter Six 186
defoliation intensity only III Year l ' , under intensive defoliation (4 cm) the
inflorescences were 32% of those recorded in 1 0 cm defoliation height treatment.
Table 6-1 6. Inflorescences/m2 (I), viable seeds/m2
(S) and 1000 seed weight (W) (g)
of BFTIWC mixture under different strategies and intensities of defoliation,
evaluated during two years.
Year 1 998-1 999 Year 1999-2000
BFT WC BFT WC
I S W I S W S W I S W
Defoliation
Strategy
S I 30 4 1 0 1 .204 \ 35 4850 0.555 20 1 50 1 . 1 1 9 2 5 0.6 1 8
S2 345 8995 1 . 1 80 2 1 5 7585 0.559 70 7 1 0 1 .2 1 1 5 35 0.6 1 6
S3 540 1 1 1 05 1 .244 345 7 1 35 0.542 1 00 805 1 .206 2 0
S4 340 7020 1 . 1 89 3 1 0 1 1 360 0.55 1 200 1 800 1 .246 6 30 0.478
SEM (n= 16) 46. 1 1 594 0.0734 96.3 1432 0.0369 29.2 300 0.0547 1 .5 9 0.0271
Signif. * * * * NS NS NS NS NS ** NS NS NS
Defoliation
Intensity
4 cm 260 4850 1 .239 1 20 4 1 00 0.553 90 765 1 . 1 82 1 0 0.53 1
1 0 cm 365 8920 1 . 1 76 3 80 1 1 365 0.55 1 1 05 965 1 .228 6 30 0.573
SEM (n=32) 32.6 1 1 27 0.0388 68. 1 1 0 1 3 0.0265 20.7 1 97 0.0325 1 .0 7 0.01 92
Signif. NS NS * * * NS NS NS NS * NS
** , P<O.O l ; P<O.05; NS, not significant; SEM, standard error of the mean; (n), number of observations for each treatment mean
The number of viable seeds produced was affected by defoliation strategy in BFT
during the two years (P<O.O l in both cases). The un spelled treatment (S I ) had the
lowest number of viable seeds, numbers increasing with winter rest in Year I or autumn
rest in Year 2. The intensity of defoliation affected the number of viable seeds (P<O.05)
in Year I , viable seeds in the intensely defoliated swards being 54% of viable seeds
produced when defoliated at 1 0 cm height.
Chapter Six 187
The number of viable seeds in WC was unaffected during Year 1 by defoliation
strategies, but there were differences (P<0.05) in Year 2 despite the low seed
production. The spell treatment increased viable seeds, and the autumn rest resulted in
further improvement (Table 6- 1 6). Defoliation intensity significantly affected viable
seeds production in the two years (P<O.O I and P<0.05 respectively), in both cases seed
numbers being greater in swards defoliated at 1 0 cm than at 4 cm.
The 1 000 seed weight of BFT and WC was not affected by defoliation treatments, and
did not differ significantly between years. General means for 1 000 seed weight were
1 .20 and 0.56 g for BFT and WC respectively.
6.4.5.2 Patterns of seed production
In 1 998- 1 999, mature BFT seed collection began on 22 January and continued during
February. Over the season, significant interactions time x defoliation strategy (P<O.O l )
and time x defoliation intensity (P<O.O 1 ) were observed (Figure 6-3 a,b). The spelled
treatments (S2, S3 and S4) started to produce mature seeds earlier than the unspelled
treatment (S I ). In early February, the winter rest treatment (S3) was significantly more
productive than the others and the unspel led treatment (S I ) the poorest seed producer.
At the last sampling in late February, defoliation strategies stil l affected seed
production, the unspel led treatment (S 1 ) producing less than the other treatments.
During the first year, the effect of defoliation intensity on BFT seed production was
observed only in the first seed collection. At that time the intensively defoliated plots (4
cm) produced only 1 4% of the seed produced by lax defoliated treatments (Figure 6-3
b)
During 1 999-2000, BFT seed production was poor compared with the previous year.
There were effects of defoliation strategies only in early January (Figure 6-3 a), when
the autumn rest (S4) produced more than the other treatments. There were no effects of
defoliation intensity during the season (Figure 6-3 b). Seed production started earlier in
WC than in BFT during the 1 998- 1 999 season. The first records of mature seeds were
on 24 December (Figure 6-4). Significant interactions time x defoliation strategy
(P<0.05) and time x defoliation intensity (P<O.O I ) were observed during 1 998- 1 999.
Chapter Six 188
There were significant effects of defoliation strategy on WC seed production in early
January and early February (P<0.05, in both cases) (Figure 6-4 a). The unspelled
treatment (S 1 ) produced less than the spelled treatment (S2). The intensity of
defoliation affected WC seed production from late December to early February (Figure
6-4 b), in all cases 1 0 cm defoliation height producing more seed than 4 cm height.
During 1 999-2000, there were no effects of defoliation strategies and defoliation
intensities, seed production was scarce and concentrated during the sampling in early
January (Figure 6-4).
1 2
29 -Dec I O -lan n-lan 8-Feb 26-Feb
Year I
1 2
N S
NS NS
29 -Dec I O -lan 2 2 -Jan 8-Feb 26-Feb
Year I
NS
I!! 7-Jan 1 4 -1an 2 1 -1an
Dates 0 f samp ling
(b)
NS NS r;s
NS
III 2 7 -Jan
Year 2
NS
RI
� S I
-o-- S2
--.:.-S3 -S4
NS
N��
I O-Feb 1 7 -Feb 24-Feb
--0--4 cm
--0-- 1 0 cm
NS NS NS NS
7-Jan 1 4 -Jan 2 1 -Jan 2 7 -Jan I O -Feb 1 7-Feb 24-Feb
Year 2
Dates o f sampling
Figure 6-3. Patterns of seed production in BFT (g/m2) over two summer seasons
affected by defoliation strategies (a) and by defoliation intensities (b). **, P<O.01 ;
* , P<O.05; NS, not significant; numbers i n brackets, SEM (na=16, nb=32).
Chapter Six
'" o
1 2
9
.� 6 -g li.
I NS 3
1 NS
(0.30)
o l _ _ �_� �-� - �-�-� _�_,� _
NS
29-Doc 1 0-Jan 22 -Jan 8 -Feb 26-Feb
Ne � '" o
1 2
� 6 ." o li. ] (0.24) '"
Year J
29-Doc I O-lan 22-J.n 8-Feb 26-Feb
Year I
(a) --<>- S I
-<>- S2 -tr-S3 - S4
NS NS
NS NS NS NS NS
7-Jan 1 4 -Jan 2 1 -ian 27-Jan I O -Feb 1 7 -Feb 24-Feb
Dates of samp ling
(b)
NS NS
Year 2
NS NS
--<>-4 cm --0- 1 0 cm
NS NS NS
189
7-Jan 1 4 -Jan 2 1 -Jan 27 -Jan I O -Feb 1 7-Feb 24-Feb Year 2
Dates ofsampling
Figure 6-4. Patterns of seed production in WC (g/m2) over two summer seasons
affected by defoliation strategies (a) and by defoliation intensities (b). **, P<O.01 ;
*, P<O.05; NS, not significant; numbers in brackets, SEM (na=16, nb=32).
6.4.6 Soil seed reserves
In April 1 998, initial soil reserves were 4340± 1 0 1 5 and 2570± 1 1 49 seeds/m2 for BFT
and WC respectively. Thousand seed weight was 1 . 1 7 1 ± 0.0 1 0 g for BFT and
0.580±0.0 1 1 g for WC, with 64% of hard seeds in BFT and 78% in WC.
Chapter Six 190
BFT soil reserves in March 1 999 were s ignificantly affected by both grazing strategies
and grazing intensity (Table 6- 1 7), being 1 0% less than in April 1 998 in S I and 59, 78
and 66% greater in treatments S2, S3 and S4. Treatment contrasts for WC were similar,
though in all treatments reserves were substantially higher in 1 999 than 1 998. Seed
reserves were greater following lax ( 1 0 cm) than severe (4 cm) defoliation in both BFT
and WC, in both cases increasing seed reserves over the initial sampling ( 1 998). In
March 1 999, BFT 1 000 seed weight was not affected by either grazing strategy or
severity, and in WC was affected only by grazing severity (Table 6- 1 7).
BFT soil reserves in March 2000 were affected by defoliation strategy (P<O.O l ) and by
defoliation intensity (P<O.O 1 ). There was an effect of spell period on seed reserves
(Table 6- 1 7), but there were no differences between spel l treatments. Intensive
defoliation (4 cm) reduced reserves compared with lax defoliation ( 1 0 cm). A
significant interaction time x defoliation height (SEM 639, P<0.05) was observed, soil
seed reserves declined from initial values when the unspeUed treatment (S 1) or
intensive defoliation (4 cm) were applied.
WC reserves in March 2000 were affected by defoliation strategy (P<0.05) and
defoliation intensity (P<O.O 1 ) (Table 6-1 7). There was no effect of spell treatment
(S 1 =S2), but the autumn rest improved seed reserves over the unspel led treatment. The
intensive defoliation (4 cm) showed lower WC reserves than lax defoliation ( 1 0 cm)
(Table 6- 1 7). Over time, there was a significant interaction time x defoliation height
(SEM 1 033 , P<0.05). WC seed reserves increased at the end of first year, then declined
but values remained higher than initial values (April 1 998) in all cases.
At the end of the evaluation in March 2000, thousand seed weight ofBFT and WC were
not affected by defoliation strategy or intensity (Table 6- 1 7). Also, there were no
significant effects over time in either BFT or WC 1 000 seed weight.
Chapter Six 191
Table 6-17. Soil seed reserves and 1000 seed weight parameters in mixed birdsfoot
trefoil (BFT) and white clover (WC) swards under different defoliation strategies
and intensities, during two years.
March 1 999 March 2000
BFT BFT wc wc BFT BFT wc wc Seeds/ml
1000 seed weight Seedsfm1 1000 seed weight Seeds/ml 1 000 seed weigh! Seedslm1 1000 seed weight
(no.l ml) (g) (no.l m:) (g) (00.l m1) (g) (no.l m1) (g)
Defoliation
strategy
S I 3895 1 .2 1 7 7670 0.551 1 8 1 0 1 . 1 76 3920 0.527
S2 6890 1 .235 8780 0.556 3965 1 .229 4550 0.567
S3 7725 1 .234 8980 0.547 3450 1 .25 1 6620 0.560
S4 7225 1 .262 14070 0.559 4520 1 .299 7495 0.556
SEM (n) 968 (48) 0.029 (8) 1 682 (48) 0.007 (8) 5 1 2 (48) 0.0337 (8) 989 (48) 0.0 1 58 (8)
Significance NS NS • • NS NS
Defoliation
Intensity
4 cm 5050 1 .258 6 1 65 0.544 2395 1 .237 3600 0.557
1 0 cm 7 8 1 5 1 .2 1 5 1 3580 0.562 4475 1 .24 1 7690 0.548
SEM (n) 684 (96) 0.021 ( 1 6) 1 1 90 (96) 0.005 ( 1 6) 361 (96) 0.0238 ( 1 6) 699 (96) 0.0 1 1 1 ( 1 6)
Significance • • NS • • • • NS • • NS
** , P<O.O 1 ; * , P<O.05; NS, not significant; SEM, standard error of the mean; (n), number of observations for each treatment mean
6.4.7 Seedling emergence
The two species followed similar emergence patterns in control led field conditions
without any sward competition between June and December 1 998 (Figure 6-5, Plate 6-
2), achieving 1 860 and 880 emerged seedl ings/m2 for BFT and WC respectively,
corresponding with 44 and 3 5% of potential emergence from the 1 998 seed bank.
During winter, there was 76 and 7 1 % of total emergence for BFT and WC respectively.
In spring, emergence was low in the two species, pulses of seedling emergence
occurring after periods of rain, combinations of rain and drought and peaks of variation
in soil temperature. However, there was a general decline in seedling emergence to the
end of spring.
Chapter Six 192
Plate 6-2. Seedling emergence was checked regularly from soil cores placed in an
adjacent area to experimental site and maintained free of ground cover.
Chapter Six 193
(a) Emerged BFT seedlings 1m2 /week
500
400
300 � �: I��lJ _ .�" � _L-" _ _ --'LJ--.ll _ _ __ n J _����""�"� � �
500 .. 400 J 300 i 200 �I
(b) Emerged WC seedlings/m2/week
100 j I I o JI_t�� - � " �� �I� I ��_�I � _ ���. � �� I_� �I ���I� � � � _��_ .��_� �t .� _.
( c ) C l i m a t i c p a r a m e t e r s
4 0 3 5 3 0 2 5 �
2 0 �
1 5 �
1 0 5 �
D a i l y S o i l T e rn p e ra tu re ( ° C )
o � ---�"". - �""� .- -- --- .��.�
2 0 - J u n 1 8 - J u 1 1 5 -A u g
W e e k ly R a i n fa l l ( m m ) r 1 5 0
I� 1 2 0
9 0
" 6 0
___ L __ _ l lLLo_Lf : o 1 2 - S e p I 0 - 0 c t 7 - N o v 5 - D e c
I�"�-�"" --�- ·"-·---""-- ·-·�·- �"-"�"· ��"· �·��-� �"-��·��""-"--"�.
L�� e e k ly R�n fa l l -- M i n i m u n te m p e ra t u re -- M a x i m u n t e m p e ra tu re
" I --.� �.�-�� .. � ... -.� .. ���� .. - .. ��-. .... �.� .... �-.-� .. �--�.-� ... �- .. ----... �--� " -
Figure 6-5. Seedling emergence patterns of (a) birdsfoot trefoil (BFT) and (b)
white clover (WC) and (c) climatic parameters, evaluated on field from June to
December 1998.
Chapter Six 194
From March to August 1 999, seedling emergence under sward competition was
substantially lower than values shown in Figure 6-5, varying from 5- 1 3% in BFT and
4-7% in WC of 1 999 seed reserves (Table 6- 1 8). BFT seedling emergence showed
significant effects of defoliation strategy, with particularly high emergence in S3,
mainly associated with a high seed input during spelling. A reduction of sward
competition by intense defoliation (4 cm) promoted an increase of 7 1 % in BFT
seedling emergence compared with more lax defoliation ( 1 0 cm). In contrast, WC did
not show any treatment effect on seedling emergence, though a high percentage of
recruitment was observed under severe defoliation for both BFT and WC (Table 6- 1 8).
Table 6-18. Seedling emergence (no.!m2) and percentage of emergence from soil
seed reserves of birdsfoot trefoil (BFT) and white clover (WC) under different
strategies and intensities of defoliation from March - August 1999.
Defoliation strategy
S I
S2
S3
S4
SEM (n)
Significance
Defoliation intensity
4 cm
1 0 cm
SEM (n)
Significance
Seedling emergence
March 1 999-August 1 999
BFT
320
4 1 5
825
460
1 26 (8)
*
640
375
89 ( 1 6)
*
(no./m2)
WC
325
420
650
640
1 20 (8)
NS
450
570
85 ( 1 6)
NS
Percentage of emergence
from March 1 999 soil seed reserves
BFT
8
6
1 1
6
1 3
5
(%)
WC
4
5
7
5
7
4
*, P<0.05; NS, not significant; SEM, standard error of the mean; (n), number of observations for each treatment mean
Chapter Six
6.5 DISCUSSION
195
The results showed a high degree of variation between years due to changing climatic
conditions. The herbage production of Year 2 was 77% of that obtained in Year 1 . The
seed production of BFT and WC in Year 2 was 1 3 % and 2% of production obtained in
Year 1 respectively. BFT plant density declined 45% during the second year and active
growing points of WC almost disappeared at the end of the second year.
Species that grow in pastoral areas of Uruguay are affected by irregular droughts and
wet periods. These conditions determine that grazing management should be oriented to
obtain a high grazing efficiency without putting at risk the survival of species
introduced into native communities. Comparatively, BFT has a high degree of tolerance
to drought conditions compared with WC, by the presence of a deeper taproot system
(Seaney and Henson, 1 970) that contrasts with the shallow root system of Wc. The
drought conditions and high soil temperatures recorded during the second summer
caused a 45% reduction in BFT stand and a massive death of WC plants. The death of
stolons in WC is affected by these factors (Belaygue et al. , 1 996; Woodfield and
Caradus, 1 996), which are reported as critical during summer in Uruguay (Canimbula,
1 977). The average soil temperature (registered at 5 cm depth in soil without vegetation
cover) was 30.9, 33 .4 and 30.4 °C for December, January and February respectively
with maximum values of 36, 37 and 34 °C for the respective months.
To increase annual herbage accumulation, the autumn rest was an effective strategy
during the two years and winter rest only in the second year. Lax grazing ( 1 0 cm)
increased the herbage accumulation between 1 8 and 2 1 % in comparison with intensive
defoliation (4 cm). BFT contribution was affected by defoliation management during
the two years, accumulation increasing by lax defoliation ( 1 0 cm) and by strategies with
extended rest periods (S4), effects that showed a similar tendency in WC during the two
years. There were no advantages to the accumulation of BFT in spring from winter rest.
The benefits produced by autumn rest were in accordance with results reported in
Chapter 3. The inclusion of an early autumn rest in BFT swards promotes herbage
Chapter Six 196
production in spring. These results suggest that BFT/WC mixtures should receive a rest
period in autumn to increase annual productivity, and this will be discussed in detail in
the final discussion (Chapter 7). Previous work at INIA Treinta y Tres with BFTIWC
oversown mixtures showed that rest periods between 60 and 80 days were enough to
enhance herbage production without excessive losses of quality (Canimbula and Ayala,
1 995) .
The advantages of autumn rest to herbage accumulation were associated with an
improvement in morphology of BFT plants. Primary plant branching of BFT increased
in the autumn rest treatment during the first year, but a decline over time occurred for
all grazing strategies. The intensive grazing (4 cm) reduced plant branching in BFT, and
differences were observed in December of both years. Also, secondary branches were
reduced by intensive grazing, as occurred in swards under close cutting (Chapter 3) .
The root mass, crown size and root diameter of plants intensively grazed (4 cm) tended
to be reduced if compared with those plants lax grazed ( 1 0 cm), tendencies that are in
agreement with those presented in Chapters 3 and 4 when a range between 2 to 1 0 cm
of defoliation were contrasted. Autumn rest strategy determined that BFT plants had
stronger crowns than in the other treatments. In addition, this was associated with
differences in root mass, S4 being more rooted than S I and S2. At the end of the
evaluation these differences disappeared, probably by a reduction in stand density. The
rest in autumn (S4) or winter (S3) increased root diameter of BFT plants in comparison
with treatments that did not receive rest (S 1 and S2).
Although frequency of defoliation was not studied in this trial, the results obtained
suggest that strategies that included grazing all year-round with monthly intervals
between grazing were excessive, adversely affecting herbage production and persistence
and differences being higher during the second year (Table 6-4, Figure 6-2). In
comparison, results reported on Chapter 3 showed that the frequency of defoliation was
not a significant factor if studied during a short period (spring in that case). Thus,
management shows a cumulative effect, suggesting that BFT can only tolerate
inappropriate defoliation for short periods. Climatic conditions exerted the strongest
influence on the final results. Bologna ( 1 996) showed a negative effect of defoliation
Chapter Six 197
intervals shorter than 4 weeks on BFT production. Several studies reported that BFT
management needs to be a compromise between frequency and intensity of defoliation,
combinations of frequent and close grazing being inappropriate (Smith and Nelson,
1 967; Greub and Wedin, 1 97 1 a; 1 97 1 b). The favourable effects of rotational grazing
rather than continuous grazing in BFT were early observed by Van Keuren and Davis
( 1 968).
A high seed yield was obtained under favourable climatic conditions, achieving an
average of 1 1 and 1 0 times the sowing rate used for BFT and WC respectively. The
spelling period increased BFT seed production 1 6 times, the winter rest further
increasing seed production. The seed production of WC was not affected by grazing
strategies. WC tolerated more intensive defoliation than BFT for seed production. Seed
production of an unspelled treatment was 5 times higher than the seeding rate used in
WC.
Conditions for flowering are controlled by daylength and temperature. In WC, higher
temperatures and longer days (> 1 2 hours) at the end of spring favour flowering (Hill et
al., 1 999), and BFT requires a minimum daylength between 1 4 to 1 4.5 hours (McKee,
1 963). In this experiment, WC showed an earlier seed production period than BFT,
related to these factors.
The summer spelling period in the experiment extended from early December to late
February. Results from Year 1 showed that seed yield between December and early
February was more than 82 and 90% of total yield in BFT and WC respectively. Despite
the indeterminate growth habit of BFT and successive fluxes of flowering, Li and Hill
( 1 988) showed that more than 70% of inflorescences are produced in a short period (25
days). In WC, seed production is not the only mechanism involved in reproduction,
vegetative reproduction being a more important alternative. Based on these arguments,
it can suggested that 60-70 days from December is a long enough spelling period for
BFTIWC mixtures. In practice, this allows a period of intensive grazing of mature
forage at the end of summer to clean swards for recruitment of new seedlings in
autumn.
Chapter Six 198
The size of a soil seed bank is the result of previous inputs by reproduction of parental
plants and eventual ly reflects the inputs by sown or dispersed seed (Pearson and Ison,
1 997). The quantification of reserves showed densities from 1 800 to 7800 seeds/m2 of
BFT and from 2500 to 1 4000 seeds/m2 of WC. Arana and Pifieiro ( 1 999) working with
WC Zapican in Uruguay, determined annual inputs to the soil seed bank between 2600
and 1 2000 seeds/m2. Bologna ( 1 996) reported soil seed reserves of BFT Grasslands
Goldie between 1 9000 to 28000 seeds/m2, in environments of South Island of New
Zealand. In the case of Lotus pedunculatus Maku, seed banks larger than 6000 seeds/m2
are required for persistent swards in high latitudes (> 32°S) of Australia (Blumenthal
and McGraw, 1 999).
Bologna ( 1 996) found 1 5% of soil seed reserves in the first 2 cm of soil strata of swards
defoliated at 2 week intervals, in contrast with 60% of those defoliated at 8 week
intervals. Differences were attributed to trampling by sheep, particularly in wet
conditions or by intensive grazing. In this experiment, the fraction of seeds under 5 cm
depth was minimal, and the probabi lity that these seeds would produce viable seedlings
was low.
The average of seedling emergence was 5 - 1 3% in BFT and 4-7% in WC of soil seed
reserves. Seedling emergence appears associated with seed inputs produced by summer
spelling. The soil seed bank can act as a buffer maintaining relative rate of emergence if
annual seed input is reduced, as occurred in intensive grazing schemes in Year 1 or
when seed input declined in Year 2 due to poor climatic conditions. Emergence was
also improved by intensive defoliation (4 cm), results that agree with those of Bologna
( 1 996) who found in BFT an increase in seedling recruitment under frequent
defoliation, recruitment being lower than 1 0% for an environment of South Island of
New Zealand. A fraction between 3 5 and 42% of seeds, depending on the species, was
activated if sward competition was eliminated, exposing seeds to fluctuating
temperature and humidity to break down dormancy.
These facts suggest that the low efficiency of the seedling recruitment process needs to
be augmented by additional management strategies especially in autumn. The intensive
defoliation reduces sward competition, creating gaps (Pearson and Ison, 1 997) for
Chapter Six 199
seedling establishment. These requirements need to be compatible with prevIOus
recommendations in terms of advantages of autumn rest to improve herbage
accumulation. It can be suggested that in those years when stand density needs to be
improved, extended autumn rest should be avoided, giving opportunity for new
seedlings establishment. Based on the potential seedling emergence results, in those
cases when plant frequency of species of interest is low and adequate densities of seed
are present in the soil , more extreme intensities of sward disturbance (eg. the
application of herbicides, intensive grazing or soil disturb) could be practised to
accelerate recruitment.
6.6 CONCLUSIONS
Evidence from this trial suggests that management to improve herbage accumulation of
BFT/WC mixtures requires a rest period in autumn, grazing swards at lax intensities
when monthly intervals are used between grazing cycles. Summer spelling for 60-70
days starting in December is necessary to increase soil seed reserves, management that
is further enhanced by autumn rest or winter rest to improve WC or BFT seed
production respectively. However, an extended seed spell ing period is recommended in
years where stand density or seed soil reserves decline. Recruitment of new individuals
from the soil seed bank has a low efficiency, demanding high seed reserves to have
more chance to incorporate new individuals or eventually the application of alternative
practices to activate the dormant soil seed bank fraction and increase the absolute values
for the recruitment of new individuals.
Chapter Six
6.7 REFERENCES
200
AL TIER, N. ( 1 997). Enfermedades del Lotus en Uruguay. (Lotus diseases in Uruguay).
INIA La Estanzuela, Uruguay. Serie Tecnica No. 93. ISBN: 9974-38-083-9. 1 6
pg.
ARANA, S. AND PINEIRO, G. ( 1 999). Deficit hidrico y manejo: su influencia en la
demografia y produccion del trebol blanco. (Water deficit and management: the
influence in the demography and production of white clover). Tesis Facultad de
Agronomia. Montevideo Uruguay.
AYALA W. AND BERMUDEZ R. ( 1992). Fertilizacion Fosfatada de Pasturas.
(Phosphatic fertilization of pastures). In Mejoramientos Extensivos en la Region
Este, pg 49-59. Resultados Experimentales 1992. INIA Treinta y Tres. Estacion
Experimental del Este. Octubre 1 992.
AYALA W. AND CARAMBULA M. ( 1 995). Mejoramientos Extensivos en la Region
Este: Manejo y Utilizacion. (Oversown pastures in the Eastern Region:
Management and Utilization). In Seminario de Actualizacion Tecnica sobre
Produccion y Manejo de Pasturas. pgs. XVlII- l a XVIII-5 . 1 7 al 1 9 de Octubre,
1 995. Tacuarembo. Uruguay.
BUHLER, D. D. HARTZLER, R. G. FORCELLA, F. ( 1 997). Implications of weed
seed bank dynamics to weed management. Weed Science 45: 329-336.
BELAYGUE, C., WERY, J., COWAN, AA AND TARDIEU, F. ( 1996). Contribution of
leaf expansion, rate of leaf appearance, and stolon branching to growth of plant
leaf area under water deficit in white cover. Crop Science 36: 1 240- 1246.
BLUMENTHAL, MJ. AND McGRA W, R.L. ( 1 999). Lotus adaptation, use and
management. Trefoil: The science and technology of Lotus. CSSA Special
Publication Number 28. Ed. by P.R. Beuselinck. American Society of
Agronomy, Inc. Crop Science Society of America, Inc. Madison, Wisconsin,
USA pp. 97- 1 1 9.
BOLOGNA, JJ. ( 1 996). Studies on strategies for perennial legume persistence in
lowland pastures. Thesis of Master of Agricultural Science. Lincoln University.
New Zealand. 220 p.
Chapter Six 201
CARAMBULA, M. ( 1977). Manejo de pasturas sembradas. (Management of cultivated
pastures). Editorial Hemisferio Sur. 374 pp.
CARAMBULA, M. AND AYALA, W. ( 1 995). Algunas Pautas de Manejo de
Mejoramientos Extensivos. (Management recommendations for oversown
pastures). In Mejoramientos Extensivos: Manejo y Utilizaci6n. Area Produccion
Animal. Serie de Actividades de Difusion No. 75. INIA Treinta y Tres. pgs. 1 2- 1 8.
Treinta y Tres. Octubre, 1 995.
CARAMBULA, M., CARRIQUIRY, E. AND AYALA, W. ( 1 994). Siembra de
Mejoramientos en Cobertura. (Establishment of oversown pastures). In Boletin de
Divulgaci6n No. 46. 1 994. INIA. ISBN: 9974-38-0 1 5-4. 20 pg. Junio 1994.
GREUB, L.J. AND WEDIN, W.F. ( 1 97 1 a). Leaf area, dry-matter accumulation, and
carbohydrate reserve levels of birdsfoot trefoil as influenced by cutting height.
Crop Science 1 1 : 734-738 .
GREUB, L.J. AND WEDIN, W.F. ( 1 97 1 b). Leaf area, dry-matter accumulation, and
carbohydrate reserves of alfalfa and birdsfoot trefoil a three-cut management.
Crop Science 1 1 : 34 1 -344.
GOERING, H.K. AND VAN SOEST, PJ. ( 1 970). Forage Fibre Analysis. U.S.D.A.
A.R.S . Agricultural Handbook No. 379.
HILL, MJ.; HAMPTON, J.G. AND ROWARTH, J.S. ( 1 999). Herbage seeds. Chapter
1 6. New Zealand Pasture and Crop Science. Edited by l White and l Hodgson.
Oxford, University Press, Auckland. ISBN 0- 1 9-558375-2. pp. 249-262.
ISTA. ( 1 985). International Rules for Seed Testing 1 985. Proceedings of the
International Seed Testing Association 13 : 299-520.
LI, Q. AND HILL, M.l ( 1 988). An examination of different shoot age groups and their
contribution to the protracted flowering pattern in birdsfoot trefoil (Lotus
corniculatus L.). Journal of Applied Seed Production 6: 54-62.
McKEE, G.W. ( 1 963). Influence of daylength on flowering and plant distribution in
birdsfoot trefoil. Crop Science 3:205-208.
PEARS ON, CJ. AND ISON, R.L. ( 1 997). Generation. Chapter 3. Agronomy of
Grasslands Systems. 2nd Edition . . Cambridge University Press. pp 37-59.
PRESTES, N.E. ( 1 995). Sobressemeadura do cornichao (Lotus corniculatus L.) cv. San
Gabriel em pastagem natural - diferemento e adubayao. (Sowing of birdsfoot
trefoil (Lotus corniculatus L.) in grass lands - stockpilling and fertilization.)
Chapter Six 202
Disserta9ao de Mestrado em Zootecnia (Plantas Forrageiras), Facultade de
Agronomia, Universidade Federal do Rio Grande do SuI, Porto Alegre, Brasil .
1 1 8 p
ROBERTS, H.A. AND BODDRELL, lE. ( 1 985) . Seed survival and seasonal pattern of
seedling emergence in some Leguminosae. Annals of Applied Biology 106: 1 25-
1 32 .
SAS INSTITUTE ( 1 990). SAS/STAT User 's Guide, Version 6. Cary, NC: SAS Institute.
SEANEY, R.R. AND HENSON, P.R. ( 1 970). Birdsfoot trefoil . Advances in Agronomy
22: 1 1 9- 1 57 .
SMITH, D. AND NELSON, C.J. ( 1 967). Growth of birdsfoot trefoil and alfalfa. 1 . Responses to height and frequency of cutting. Crop Science 7: 1 30- 1 33 .
TILLEY, lM.A. AND TERRY, R.A. ( 1 963). A two-stage technique for the i n vitro
digestion of forage crops. Journal of the British Grassland Society 18: 1 04- 1 1 1 .
VAN KEUREN, R.W. AND DAVIS, R.R. ( 1 968). Persistence of birdsfoot trefoil,
Lotus corniculatus L. as influenced by plant growth habit and grazing
management. Agronomy Journal 60: 92-95 .
WAGHORN, G.C . AND SHELTON, LD. ( 1 997). Effect of condensed tannins in Lotus
corniculatus on the nutritive value of pasture for sheep. Journal of Agricultural
Science, Cambridge 128: 365-372.
WOODFIELD, D.R. AND CARADUS, J.R. ( 1 996). Factors influencing white clover
persistence in New Zealand pastures. Proceedings of the New Zealand
Grassland Association 58: 229-235 .
Chapter Seven
7. INTEGRATING DISCUSSION
7. 1 INTRODUCTION
7.2 BIRDSFOOT TREFOIL GENOTYPES
203
7.3 DEFOLIATION MANAGEMENT, PRODUCTION AND PLANT
SURVIVAL
7.4 THE ROLE OF THE SOIL SEED BANK ON POPULA TION
DYNAMICS AND PERSISTENCE OF BIRDSFOOT TREFOIL
7.5 PRACTICAL MANAGEMENT RECOMMENDATIONS
7.5 . 1 Seasonal management
7 .5 .2 General management
7.6 CONCLUSIONS
7.7 REFERENCES
Chapter Seven
7.1 INTRODUCTION
204
Birdsfoot trefoil (BFT) is recognised as a valuable feed source in many areas around the
world, with special contribution in marginal areas, where l imitations in tolerance of low
ferti l ity, low pH or drought limit the productivity of other legumes commonly used.
Two major reasons for renewing interest in BFT nowadays are its value for low input
systems and its high feed value particularly due to the presence of condensed tannins
(Section 2.2.4.5). Despite extended research over decades, the weakness of BFT under
intensive grazing and the incidence of crown-rot diseases are the main unsolved
problems that limit long term persistence.
The development of controlled management strategies to improve productive
persistence ofBFT in pastoral systems of New Zealand and Uruguay provided the focus
for this work. A series of four field and glasshouse experiments (Table 7- 1 ) were
conducted in Palmerston North, New Zealand (latitude 40°23 ' S) and Treinta y Tres,
Uruguay (latitude 33°54' S) with the objective to determine appropriate defoliation
strategies for different BFT cultivars, quantifYing morphological and physiological
plant adaptations under defoliation and analysing population dynamics and strategies to
improve BFT persistence. Despite the importance attributable to disease incidence, this
was not a specific objective in these experiments.
These objectives were explored over four BFT cultivars with contrasting plant structure.
Defoliation strategies included a range of intensities between 2 to 1 0 cm and 4 to 1 0 cm
under cutting and grazing, respectively (Table 7- 1 ). Defoliation intervals of 20 and 40
days were contrasted in spring, and 20, 30 or 40 day intervals were applied in those
cases when this was not studied as a variable. Timing of initial cutting in the year of
establishment varied from vegetative to late mature stages, and swards evaluated varied
from one to four years old, pure BFT (Experiment 1 to 3) or mixed with white clover
and native grasses (Experiment 4).
Chapter Seven
Table 7-1. Description of experiments conducted in this project.
Experiment Location
I . (Chapter 3) DRU, Massey University,
New Zealand
2. (Chapter 4) PGU, Massey University,
3 . (Chapter 5)
4. (Chapter 6)
New Zealand
Palo a Pique, INIA,
Uruguay
Palo a Pique, INIA,
Uruguay
Defoliation variables
Intensity (2, 6 and 1 0 cm)
Frequency (20 and 40 days)
with or without autumn rest
Intensity (2, 6 and 1 0 cm)
Frequency (20 days)
Intensity (4 and 8 cm)
Frequency (40 days)
Timing
(vegetative, flowering ,
maturity)
Cultivars (4)
Intensity (4 and 8 cm)
Frequency (30 days)
Grazing strategies (4)
Pasture type/age
Pure, 3 years
Plants, 3 years
Pure, 1 -2 years
Mixed, 3-4 years
205
Period
April 1 997 - December 1 997
September 1 997 - December 1 997
May 1 998- April 2000
April 1 998- March 2000
DRU, Deer Research Unit; PGU, Plant Growth Unit; INIA, National Institute of Agricultural Research
The integrating discussion, which follows, is structured in sections:
(i) Evaluation of the characteristics, adaptabil ity and performance of the four
genotypes studied and the relevance of this information to future genotype
developments.
(ii) Analysis of the effect of defoliation management on herbage production and
plant survival, with particular reference to the intensity (main factor studied),
frequency and timing of defoliation.
(iii) Assessment of the factors affecting sward persistence, with particular reference
to elements of seed production and seedling recruitment which contribute to the
maintenance of plant population density.
(iv) Development of a series of practical recommendations for seasonal and general
management of BFT swards, drawing together the conclusions from the
preceding sections.
Chapter Seven
7.2 BIRDSFOOT TREFOIL GENOTYPES
206
A wide range of variation in behaviour and production was observed between genotypes
studied. Plant types varied from semi-erect and erect (San Gabriel and INIA Draco) to
semi-prostrate types (Grasslands Goldie and Steadfast). Morphological parameters and
biomass distribution varied over the experiments, genotypes reacting differently to
defoliation. Shoot density is a desirable character in BFT to improve herbage production
(Figure 3 -5), intensive defoliation reducing the number of shoots per plant. Grasslands
Goldie showed a higher density of primary and secondary shoots than the other
genotypes (Section 5 .4.8). The production of shoots could be associated with crown
size, under the hypothesis that bigger crowns could develop more new shoots. However,
the effects of defoliation on the rate of replacement of shoots are stil l unclear and should
be more exhaustively studied.
The general effects of defoliation on the rate of replacement of intensively defoliated
BFT plants (Tables 3-4 and 4-4, and Sections 5 .4 .8 and 6.4.4), suggest than those plants
with a low ratio of abovelbelow-ground biomass, a strong root system and a bigger
crown could be more tolerant of severe defoliation, with the risk to be less productive.
The role of root reserves could be enhanced in plants with wel1 -developed root systems.
The selection of genotypes with these improved characteristics could help to overcome
persistence problems (Nora Altier, personal communication). However, the results
showed that the performance of cultivars was primarily l imited by environmental rather
than morphological constraints.
Herbage production showed a wide range of variation between cultivars when compared
in Uruguay, annual production of local cultivars being more than two times higher than
introduced cultivars (Figure 5-2). The degree of winter activity of BFT genotypes at low
latitudes (Uruguay) resulted in important differences in herbage production. New
Zealand latitudes range from 33° to 47° S approximately, in contrast with Uruguay with
latitudes between 30° to 35° S. Genotypes currently used in Uruguay (San Gabrie l and
INIA Draco) showed a degree of activity in winter and early spring that contributed to
substantial advantages in production and periods of utilisation in the year over
Chapter Seven 207
introduced cultivars. Winter dormant genotypes, in which production is concentrated
from end of spring to late summer, are not recommended at low latitudes due to a short
growing season. In New Zealand, there is available only one commercial BFT cultivar
(Grasslands Goldie), adapted to grow at higher latitudes and under more extreme winter
conditions like those occurring in environments of the South Island of New Zealand.
However, in the North Island of New Zealand, at lower latitudes and under warmer
conditions, there is scope to evaluate the potential of winter active genotypes. Despite
the limited area of BFT sown in New Zealand, the study of other BFT cultivars adapted
to specific environments could contribute to increase the interest in this species.
The novel development of a rhizomatous characteristic in cv. Steadfast offers
opportunity to shortcut some of the reported problems that affect plant persistence in
crown-forming plants, as has recently been demonstrated in red clover (Hyslop et al.,
1 999). Steadfast showed a low potential of productivity (Figure 5-2, Table 5-6),
probably due to the winter dormant characteristic, and the rhizomatous character is not
an attribute of all individuals in the population. Nevertheless, breeding programmes
could introduce the rhizome characteristic in other genotypes, like winter active types,
opening more opportunities for BFT survival. The use of prostrate material, offering a
certain degree of adaptation to intensive defoliation (Figure 5-8 c, d), requires more
extensive evaluation with focus on the development of genotypes more tolerant to
intensive grazing.
The options of BFT cultivars available for Uruguay cover reasonably well the
requirements of adaptability, productivity and persistence. The breeding programme
conducted by INIA is focusing on aspects of disease resistance and improved
morphological characteristics at the same time, as was demonstrated with the recent
release of INIA Draco.
In conclusion, the results obtained in this area are providing evidence about the
importance of selecting genotypes for specific environments. More detailed information
about the production, nutritive value and morphology of alternative cultivars is required,
in particular for cultivars of recent release like INIA Draco and Steadfast where
available information is l imited.
Chapter Seven 208
7.3 DEFOLIATION MANAGEMENT, PRODUCTION AND PLANT
SURVIVAL
The intensity, frequency and seasonal timing of defoliation are considered as the main
factors involved in the development of defoliation strategies of BFT. The main
emphasis in the current study was on the defoliation height and the timing of
defoliation, defoliation interval being only partially explored because recent studies in
New Zealand on the cultivar Grasslands Goldie have increased and clarified the
information available (Bologna, 1 996).
Extensive research over the last fifty years on the intensity of defoliation (Table 7-2)
has shown that the effects of defoliation intensity are related to growth habit, upright
plant types being more sensitive to intensive defoliation (Pierre and Jackobs, 1 953) .
The available information in some cases showed advantages in the amount of herbage
harvested when more intensive defoliation was applied (Table 7-2), in particular under
extended periods of accumulation between defoliations (Cordeiro de Araujo and
Jacques, 1 974). However, productive persistence can not be maintained over time under
intensive defoliation, production and stand density being reduced when an intensive
defoliation strategy is applied for long periods of time. Thus, the general consensus is
that BFT can be defoliated frequently but not at high intensifies (Alison and Hoveland,
1 989). The relatively greater importance of residual leaf area than root reserves for
regrowth supports these recommendations (Alison and Hoveland, 1 989).
Chapter Seven 209
Table 7-2. Summary of published research with emphasis in defoliation intensity
on birdsfoot trefoil.
Authors
Pierre and Jackobs, 1 953
Duell and Gausman, 1 95 7
Twanley, 1 968
Smith and Nelson, 1 967
Greub and Wedin, 1 97 1
Cordeiro de Araujo and Jacques, 1 974
Alison and Hoveland, 1 989
Defoliation treatments Herbage production results
2.5, 5 and 1 0 cm height Close, frequent and late fal l defoliation reduced BFT yield. Cultivars responded differentially
2.5, 7.5 cm height Herbage yield was 2.5>7.5 cm in Year I , 20-day intervals differences being reduced in Year 2 Pre-bloom to seed dehiscence stages
5, 1 5 cm height Herbage yield was greater for 5 cm height
2.5, 7.6 and 1 5 .2 cm height Herbage yield in Year 1 was 2.5>7.5> 1 5.2 3 to 6 times of defoliation during cm height, excepting 5 and 6 cuts growing season treatments. During second year a higher
stubble was needed to maintain yield independently of frequency
3.8, 7.6 and 1 1 .4 cm height Herbage yield of7.6 and 1 1 .4 cm > 3.8 cm
3, 6 cm height Defoliation at more mature stages vegetative to flowering stages increased yield.
Herbage yield of 3 cm > 6 cm if only 1 cut is applied, but 6 cm >3 cm if plants defoliated more than once
5 and 1 0 cm height Harvests at intervals of 21 days and at 3 2 1 , 28 and 42-day intervals cm height reduced herbage yield
drastically 3, 5 and 1 0 cm height 2 1 -days interval
Results observed in the current project showed that in three of four studies, herbage
production was greater at lax (6- 1 0 cm) than hard (2-4 cm) defoliation (Figure 7- 1 ).
Reductions in herbage production and plant survival are the main consequences of
intensive defoliation (Table 3-2, Figure 3- 1 , Table 4-2), for a semi-prostrate cultivar,
suggesting a priori a strong effect for erect BFT types. There was a general decline in
below-ground BFT plant parameters, resulting in a reduction in shoot density per plant
and consequently in herbage yield (Figure 3-5). The reductions observed in root number
and mass indicate that BFT plants subjected to intensive defoliation are limited in the
capacity to capture resources (nutrients and water) for successful regrowth. Despite the
recognised good production in summer and tolerance to drought conditions of BFT,
these factors have influence in the survival of BFT in summer, as was confirmed by
results obtained (Figure 6-2). Despite the short time nature of Experiments 1 and 2, and
Chapter Seven 210
acknowledging this as a limitation from which to draw conclusions, the results were
consistent in showing that plant survival and production declined quickly. These results,
which were subsequently confirmed under grazing conditions (Chapter 6), emphasise
the risks of defoliation below 4 cm irrespective of defoliation frequency.
"'"' � '-' ..c: � 0 5h (l)
.::: 0; -.; ez::
(c)
(a) 100
75
50
25 __ Chapter 3 -tr- Chapter 4
0 2 6 1 0
Defoliation height (cm)
(b)
100
D::faiatim height (cm)
Year 1 Year 2
� I: j � � 50
1' / / .� / � 25 1 -=-:� ___ S2
o ��S4 T --- -r �--r-- ��
4 10 4 1 0
Defoliation height (cm)
Figure 7-1 . Results related to the effect defoliation intensity in BFT growth from
information presented in (a) Chapters 3 and 4, (b) Chapter 5 and (c) Chapter 6.
Chapter Seven 211
In Experiment 3 herbage production increased by intensive defoliation (4 cm) if
compared with more lax defoliation (8 cm) (Figure 7- 1 b). Similar results were
observed by Duell and Gausman ( 1 957), Smith and Nelson ( 1 967), Twanley ( 1 967) and
Cordeiro de Araujo and Jacques ( 1 974), questioning previous evidence on the
importance of lax defoliation. However, these results may be explained as fol lows.
First, in all these studies, the interval of defoliation was relatively long (40 days), and/or
swards received an extended rest from autumn to spring allowing plants to accumulate
reserves and re-structure root systems for the next growing season. Secondly, it is
apparent from Figures 5-7 and 5-8 (Section 5 .4 .5 .2 , Chapter 5 ), that there is a
concentration of plant dry matter (mainly stems and shoots) in the low strata of the
sward canopy in both erect and prostrate genotypes, but that almost all the entire leaf is
carried higher in the canopy than 8 cm from soil level . Thus, evidence from plant
structure suggests that residual leaf area is likely to be relatively insensitive to variation
in height of defoliation. This suggestion is reinforced by the evidence from Experiment
3 (Table 5 -8), where the amount of leaf remaining after cutting (Table 5 -8) was higher
for 8 cm than for 4 cm height, but the absolute values could be considered too low to
promote a fast regrowth (0. 1 3 and 0.43 of LAI for upright and semi-prostrate types
respectively). In conclusion, these results demonstrate that the advantages of intensive
defoliation (4 cm) are reduced over time or, as reported in other cases, lax defoliation (8
cm) is preferable if long-term experiments are analysed.
The impact of intensity of defoliation on BFT production is clearly influenced by
defoliation interval, and by the duration of the treatment. For example, measurements
over a ful l year demonstrated that when a monthly sequence of defoliation was applied,
BFT declined in production, and advantages of lax defoliation ( 1 0 cm) increased over
time (Year 2) (Figure 6-4). Rest periods in autumn or winter increased herbage
production (Figure 7- 1 c). Thus, the intensity of defoliation should be considered in
association with defoliation intervals, in terms of defined grazing cycles. Defoliation
intervals should be adjusted in accordance with seasonal patterns of growth. During
active growth in spring, there were no advantages of extended (40 days) over short (20
days) intervals . The combination lax defoliation-extended interval was less productive
than more intensive combinations in a short term experiment (Chapter 3) because of the
increase in dry matter losses in the lax defoliation treatment.
Chapter Seven 212
Long term evaluations show that defoliation intervals between 30-42 days have
advantages over short intervals (Bologna, 1 996) when swards are defoliated to 4 cm
height. When intensities of 4 and 1 0 cm where compared under 30-day intervals the
advantages to the lax defoliation became more appreciable during the second year of
defoliation (Figure 6-4), differences between intensities increasing for systems that
received between 9 and 1 2 grazing periods in contrast with those that received 6 grazing
periods.
The results of Experiments 1 and 2 (Sections 3 .4.4 and 4.4.4), suggested that shoot
population per plant was greater in plants defoliated to 6 cm than in those defoliated to
height of 2 or 1 0 cm. As in lucerne (Keoghan, 1 970), shoot population has been shown
to have a dominant influence on plant growth (Figure 3 -5) . After defoliation, plants may
compensate for the amount of herbage removed by increasing relative growth and
developing new growth sites (stems), processes that have only a short term significance
in BFT (Chapter 4). Intensive defoliation repeated over time will deplete the potential
for plants to produce new shoots, but conversely lax defoliation will reduce the potential
for development of new primary shoots by enhancing secondary shoots development.
The finding that 6 cm stubble can increase shoot numbers per plant could be explained
if defoliation is severe enough to promote the development of new primary shoots, and
stubble height is enough to contribute with sites for the development of new secondary
shoots if apical dominance is broken by cutting or grazing.
The practical significance of these effects will be influenced by the relative importance
attached to concepts of production and plant survival. Altier ( 1 997) observed that
individual BFT plants can survive for only 2-3 years. BFT swards defoliated frequently
(20 days) and close (2-4 cm) can not persist for long (Figure 3 - 1 ), and also the strategy
S I (grazing all year each 30 days) adversely affected productive parameters and plant
survival. Persistent pastures can be achieved if adequate rest periods are allowed, but
this may imply losses of production and quality. In extensive and low input systems,
attention may be focused on "productive persistence" as a major objective, and
depression in herbage production and utilisation may be acceptable consequences.
Chapter Seven 213
The timing of defoliation in the year of establishment did not affect persistence of BFT,
but there were consequences to the amount and quality of herbage harvested. Delay in
initial defoliation increased herbage accumulation, in accordance with previous reports
(Table 7-2). However, herbage quality declined with the extension of accumulation
(Tables 5 - 1 0 and 5 - 1 1 ). Digestible organic matter harvested did not increase after the
flowering stage (mid-December) when a cultivar with an early spring growth was used
(San Gabriel), but increased to advanced maturity (late January) for a later spring
growth cultivar (INIA Draco). There were no substantial losses in quality under
extended stockpiling (Tables 5 - 10 and 5 - 1 1 ), conferring more flexibility of management
opportunity. The nutritive value of BFT, analysed for different management strategies,
seasons and cultivars was in general good, reinforcing the high feeding value ascribed
to BFT (Formoso, 1 993).
In general, decline in BFT population was observed in the short term (Figure 3 - 1 ) under
close defoliation, but less intensive defoliation (8- 1 0 cm) required more time to express
the same pattern (Figure 6-2 b). Unfortunately the decline in plant density as a result of
drought conditions reduced the significance of the long term comparisons on
Experiments 3 and 4 (Figures 5-9 and 6-2). It must be assumed that the combined effect
of grazing and disease incidence will increase over time, resulting in an inevitable
decline in stand density. Thus, the philosophy of management is how to model the
"assumed and inevitable" decline in stand density (losses/outputs) with the potential
inputs or gains of new individuals, and the defoliation intensity defining the rate of
stand decline. These concepts provide the link to the fol lowing section, in terms of
identifying the best alternatives to provide natural accessions of new plants to help to
maintain a productive popUlation.
7.4 THE ROLE OF THE SOIL SEED BANK ON POPULATION DYNAMICS
AND PERSISTENCE OF BIRDSFOOT TREFOIL
The short-lived nature of BFT plants, early described by Pierre and lackobs ( 1 953), was
confirmed in the four experiments of this project for different genotypes, plant ages,
management conditions and environments, and indicate the need to develop strategies to
Chapter Seven 214
replace the losses of individual plants in BFT swards. Stand decline is not only
attributable to management practices (Table 3 - 1 ) and disease incidence (Altier, 1 997).
Adverse climatic conditions (Figure 5-9 and Figure 6-2) can produce drastic reductions
in plant population density, in some cases independently of the most conservative
management applied. Actually, the expectations of enhanced plant survival by breeding
are no longer than three or four years (Rebuffo and Altier, 1 996). The soil seed bank is
the primary and only source of plant regeneration for traditional BFT types, before
introducing pasture renewal alternatives. Understanding seed bank dynamics can help to
define management strategies on BFT swards.
Results achieved in birdsfoot trefoil/white clover mixtures showed maXImum seed
inputs of 1 1 1 00 viable seeds/m2 of BFT fol lowing summer spelling, declining under
drought conditions to less than 1 800 seeds/m2 of BFT. These disparities in seed inputs
reinforce the importance of the buffer role of the seed bank for the maintenance of plant
populations. BFT seed inputs are improved by summer spel l ing, and winter rest or lax
grazing. The reproductive structures in BFT are disposed at the top of the sward, being
easily eliminated by grazing, thus the spelling period is necessary to allow a complete
sequence from flowering to seed maturity.
Based on the climatic variation reported (Table 5-3, Figure 5 - 1 ), the development of the
soil seed bank should be promoted from early pasture stages, probably from the year of
establishment, to reduce risks of population loss under adverse climate conditions. The
soil seed reserves are variable, depending on the success of summer spelling over years.
Seed reserves between 4000 to 7500 seeds/m2, after a year of Iow seed input, were
observed in swards four years old (Table 6- 1 7). Despite the large values reported, soil
seed dynamics are quite complex, and not all seeds are in condition to produce seedlings
immediately because of dormancy mechanisms (Harper, 1 977; Pearson and Ison, 1 997).
Preliminary results from this trial (Table 6- 1 8), as well as other studies on legume
seedling recruitment (Miller et al. , 1 964; Bologna, 1 996 and Arana and Pifieiro, 1 999
among others) confirm the low efficiency of the emergence-recruitment processes. Low
rates of seedling survival are also reported in the establishment of oversown legumes,
sward competition, nodulation failures and low N2 fixation being described as the most
limiting factors (Lowther et al., 1 989), as well as climate conditions (Fraser et aI. ,
Chapter Seven 215
1 994). Additionally, low seedling vigour in BFT is a characteristic that limits rapid
establishment (Twanley, 1 967), but can be improved by breeding (Twanley, 1 967;
Frame et al. , 1 998). BFT seed weight was not altered by defoliation management (Table
6- 1 6), but an increase in seed weight is achieved by early closing (Bologna, 1 996).
Under sward competition, recruitment from the soil seed bank is frequently lower than
1 0% (Canimbula et al., 1 994), a situation that demands management practices to
increase seedling recruitment and the maintenance of a high soil seed bank to increase
potential recruitment in absolute values. More detailed information is required for the
conditions of Uruguay about patterns of seedling emergence and survival . Low winter
temperatures may inhibit seedling survival, but on the other hand can contribute to
break down dormancy of hard seeds. However, seedlings that emerge in spring have a
limited survival in summer under drought conditions that can affect new plants with an
undeveloped root system, as occurred in environments of South Island, New Zealand
(Bologna, 1 996).
There is a partial understanding of the soil seed bank dynamics and natural re seeding
processes for oversown birdsfoot trefoil/white clover mixtures. Knowledge of processes
that occur in the soil seed bank, and patterns of seedling emergence and seedling
survival, will contribute to determine strategies for spel ling frequency between years,
required seed reserves and management to increase the efficiency of recruitment. On the
mixed swards studied (Chapter 6), the manipulation of spelling processes could give
opportunities to manipulate the balance between species of interest. The percentages of
potential seedling emergence (35-44%) over a period of approximately 6 months, under
reduced sward competition, demonstrated the potential of the soil seed bank for pasture
renewal in low density stands. If there is a reasonable density of seeds in soil reserves of
species of interest, reductions of competition by grazing, application of herbicides, soil
disturbance or eventually fire could promote the emergence of desirable seedlings
increasing stand density.
In conclusion, the levels of seed production in BFT reported in this study are enough to
develop large soil seed reserves. However, the establishment of new plants was limited
by a low efficiency of seedling recruitment. These results focus the discussion about the
Chapter Seven 216
value of a seed bank for plant recruitment to maintain sward productivity and sward
persistence. In fact, the management of a BFT stand to increase soil seed reserves by
summer spelling is in conflict with the requirements to achieve the maximum
productivity and efficiency of utilisation in a short period of time. The emerging
questions of this strategy relate to the balance between the economic benefits of
increasing pasture life by adjustment of grazing to enhance natural recruitment, or
conversely establishing more intensive systems where pasture renewal is considered as
a component of the management package. Results obtained were relevant in terms of
the provision of information for extensive systems of Uruguay. The available
information in this area is limited (Olmos, 1 996; Arana and Pifieiro, 1 999), and this
thesis is basically one of the first reports to quantify reproductive processes for birdsfoot
trefoil/white clover mixtures in Uruguay. This important issue was discussed in Chapter
6, and will not be considered further here.
Clarification of these options requires more intensive research. The dynamics of soil
seed reserves, patterns of emergence, and seed inputs and outputs are crucial to
elaborate strategies to extend the productive life of BFT swards. In addition,
reproduction and dynamics of BFT plants under grazing need to be quantified to ensure
that recruitment processes constitute a viable option to maintain replacement rates of
plants . The developing knowledge in this area will contribute to the production of
predictive models of plant and seed dynamics (Emery et al. , 1 999), including the effects
of environmental factors and grazing.
7.5 PRACTICAL MANAGEMENT RECOMMENDATIONS
Based on available knowledge of BFT management (Chapter 2) and findings reported in
this work (Chapter 3 to 6), a series of recommendations can be formulated to increase
the productive persistence of birds foot trefoil pastures.
Chapter Seven 217
7.5.1 Seasonal management
Autumn
Autumn is considered a critical time for BFT, because many of the decisions made in
autumn can have a carryover effect on production and survival in the fol lowing seasons,
particularly in winter and spring. Autumn is the time where BFT plants rebuild root
reserves (Nelson and Smith, 1 968) which contribute to improved winter plant survival
and early spring regrowth. Under the conditions evaluated, winter survival was not
affected by autumn defoliation (Figure 3 - 1 ), probably because winter temperatures in
the experimental sites were not so low as to affect plant survival. However, improved
spring regrowth was observed ifBFT received an early autumn rest (Table 3-2).
In autumn, BFT plants develop new shoots from the crown, and these contribute to
spring growth (Bologna, 1 996), as occurred with BFT plants that received an autumn
rest in Year 1 of management of the BFT/WC mixture in Experiment 4 (Tables 6- 1 0
and 6- 1 1 ) . The deferment of defoliation of BFT from early to mid-autumn constitutes an
adequate practice in terms of the transfer of herbage of high quality to winter, without
effects on the persistence of BFT in BFT/WC mixtures. When stand density is reduced,
intensive defoliation in early autumn will improve the recruitment of new BFT plants
from the soil seed bank.
Recommendation:
Management of established EFT stands should avoid intensive (4 cm) and late
defoliation in autumn (June).
Winter
Winter management will be influenced by the type of cultivars in use and the place of
use. For Uruguay and when BFT is used in mixtures with white clover, winter rest
showed advantages for annual herbage production of BFT and seed production in the
following summer. When BFT is used as a pure stand, the recommendation is for
util isation in spring-summer, with a rest during winter. This scheme is fol lowed in New
Chapter Seven 218
Zealand for winter dormant cultivars. Findings on winter active lucerne cultivars in
New Zealand showed that despite the advantages in winter yield of these materials,
winter grazing resulted in significant reductions in spring growth compared with winter
dormant types (White and Lucas, 1 990). For winter active cultivars growing at low
latitudes (Uruguay), the avoidance of winter grazing did not show as great advantages
as those resulting from autumn rest.
Recommendation:
Rest periods in winter will allow increased annual herbage production and seed yields
of BFT swards, and reduce the effects of intensity of defoliation.
Spring
During the active growth period in spring, it is not recommended to use long defoliation
intervals. Otherwise, if extended intervals (40 days) and lax intensities ( 1 0 cm) are
combined, herbage losses will increase, particularly in low sward strata (Chapter 3), and
a short grazing season will result (approximately two grazing periods in the season).
Short intervals (20- 30 days) demonstrated advantages in herbage production if lax
defoliation ( 1 0 cm) was applied. In those swards that received a previous autumn rest,
early and high spring growth could be expected, in comparison with swards more
intensively grazed in autumn.
Recommendation:
A range between 6-10 cm defoliation height in combination with 20-30 days defoliation
intervals should achieve high growth rates, high efficiency of utilisation of herbage
produced, avoiding herbage losses and without risks in BFT persistence.
Summer
BFT can be managed under two contrasting criteria in this season. First, dense and pure
BFT swards can contribute with a high production because summer is a period of active
growth and adequate levels of forage quality. In those regions where BFT has a
Chapter Seven 219
restricted grazing season (spring-summer), swards have an extended rest period of
around 6 months in the year, thus plants can replenish adequately root reserves and
below-ground biomass for the next growing season and are more tolerant of defoliation
In summer.
However, in those environments where a year-round defoliation is practised, in mixed
swards and relatively old or less dense stands, a period of 60-70 days for seed spelling
starting in early December is a recommended alternative to increase persistence. Also,
lax defoliation increases summer herbage production.
Recommendation:
The maintenance of BFT swards in summer should be associated with moderate
defoliation intensifies (;? 6 cm) or even the application of res t periods to encourage seed
production and to promote seed bank reserves for the maintenance and stability of
stand.
7.5.2 General management
The seasonal recommendations formulated above need to be integrated and prioritised
when an annual grazing plan is defined for BFT. In this context, defoliation
management should avoid late autumn grazing (May-June) to encourage plants to build
root reserves and shoot development. In winter, defoliation can be practised for winter
active cultivars without excessive risks in sward persistence, particularly when BFT is
in mixtures. Frequent (20 days) and moderate intensities (6-10 cm) are proposed for
spring to achieve adequate utilisation. In the rest of the year, frequency should be
between 30-40 days approximately. For summer, management should be adjusted in
response to stand density and soil seed reserves. In the case of poor stands or reduced
soil seed reserves a spelling period for 60-70 days from early December is
recommended.
Chapter Seven
7.6 CONCLUSIONS
220
The most important conclusions from the current research program about defoliation
management to improve productivity and persistence of birdsfoot trefoil cultivars are
summarised as fol lows, providing suggestions for future research in areas where this is
required.
i) Intensive defoliation should be avoided, because it will result in a decline in
herbage production and plant survival even over short periods of time. The
reduction in shoot density by intense defoliation contributes to reductions in
herbage production, root mass and crown size. A more detailed understanding is
required of the physiological mechanisms and genetic variability involved in the
production of new shoots, and the influence of defoliation on the development of
primary and secondary shoots.
ii) BFT plants have a limited and short term plasticity in response to defoliation,
observed by the increase in relative growth rate, leaf area ratio, specific leaf area
and number of leaves per plant.
iii) Autumn rest of BFT swards contributes to improve herbage and seed
production, and may also influence plant survival in more extreme winter
conditions.
iv) In spring, defoliation management should be based on intensities between 6- 1 0
cm height and 20-30 days intervals.
v) There is a high degree of morphological variation between BFT cultivars. The
evidence suggests that crown size and root mass could be variables to be used in
breeding programmes to select cultivars with improved persistence. The
presence of rhizomes is a desirable character in BFT to increase persistence, but
Chapter Seven 221
it would be valuable to introduce this character in winter active BFT cultivars in
order to combine improved herbage production potential and stand survival.
vi) Levels of seed production in BFT are not generally l imiting to the development
of adequate soil seed reserves. Seed spelling in summer and a winter rest
improved seed BFT yields.
vii) The l imited efficiency in seedling emergence under grazing conditions raises
questions about the real value of the soil seed bank. These findings indicate the
need for increased research in this area, to develop strategies to promote a more
effective recruitment process.
viii) BFT has a recognised place among legume options for direct grazing in pastoral
systems of Uruguay. Its relevance is based on extended adaptabil ity to soil
conditions and tolerance of low soil phosphate status. However, the additional
advantages that BFT herbage confers to animal production by the presence of
condensed tannins are not fully understood and applied. The value of BFT in
pure swards or in mixtures with white clover, achieving advantages in herbage
uti l isation, bloat control and anthelmintic effects, justifY further studies on
management strategies.
ix) In New Zealand, BFT constitutes an alternative legume species, and the only
BFT cultivar is a winter dormant type with a defined grazing season during
spring�summer. The limited tolerance to intensive defoliation and competitive
capacity limit the grazing season, in comparison with other alternative species.
BFT is grazed for around 6 months of the year, with the remaining 6 months
under rest. The reduced period of uti l isation and poor tolerance to intensive
grazing and poor herbage production in comparison with traditional species,
means that farmers do not find BFT a good option to increase profitability. The
development of cultivars more competitive and tolerant to grazing could
improve interest in BFT for New Zealand farming. The potential of winter active
cultivars in some environments should be tested.
Chapter Seven 222
7.7 REFERENCES
ALISON, M.W. AND HOVELAND, C.S. ( 1 989 b). Root and herbage growth response
of birds foot trefoil entries to subsoil acidity. Agronomy Journal S1 : 677-680.
AL TIER, N. ( 1997). Enferrnedades del Lotus en Uruguay. (Lotus diseases in Uruguay).
INIA La Estanzuela, Uruguay. Serie Tecnica No. 93. ISBN: 9974-38-083-9. 1 6
pg.
ARANA, S. AND PINEIRO, G. ( 1 999). Deficit hidrico y manejo: su influencia en la
demografia y produccion del trebol blanco. (Water deficit and management: the
influence in the demography and production of white clover). Tesis Facultad de
Agronomia. Montevideo Uruguay.
BOLOGNA, J.J. ( 1 996). Studies on strategies for perennial legume persistence in
lowland pastures. Thesis of Master of Agricultural Science. Lincoln University.
New Zealand. 220 p
CARAMBULA, M., CARRIQUIRY, E. AND AYALA, W. ( 1 994). Siembra de
Mejoramientos en Cobertura. (Establishment of oversown pastures). In Boletin de
Divulgaci6n No. 46. 1 994. INIA. ISBN: 9974-38-0 1 5-4. 20 pg. Junio 1 994.
CORDEIRO DE ARAUJO, J. AND JACQUES, A. ( 1 974). Caracteristicas morfologicas
e producao de materia seca de comichao (Lotus corniculatus L.) colhido em
diferentes estadios de crescimento e a duas alturas de corte. (Morphology and
production of dry matter of birdsfoot trefoil (Lotus corniculatus L.) at different
growth stages and under two defoliation heights). Revista da Sociedade
Brasileira de Zootecnia. Vol. 3(2): 1 38- 1 47 .
DUELL, R.W. AND GAUSMAN, H.W. ( 1 957). The effect of differential cutting on the
yield, persistence, protein and mineral content of birds foot trefoil. Agronomy
Journal 49: 3 1 8-3 1 9.
EMERY, K.M., BEUSELINCK, P. , AND ENGLISH, J.T. ( 1 999). Evaluation of the
population dynamics of the forage legume Lotus corniculatus using matrix
population models. New Phytologist 144 : 549-560.
Chapter Seven 223
FORMOSO, F. ( 1 993). Lotus corniculatus. I. Performance forrajera y caracteristicas
agronomicas asociadas. (Productive performance and agronomic characteristics).
Serie Tecnica No. 37. INIA Uruguay. ISBN: 9974-556-69-4 20 pg.
FRAME, J., CHARLTON, J.F.L. AND LAIDLAW, A.S. ( 1 998). B irdsfoot Trefoil and
Greater Lotus. Chapter 6. In Temperate Forage Legumes. CAB International,
Wallingford. ISBN 0-85- 1 99-2 1 4-5. pp. 245-27 1 .
FRASER, W.J., OGDEN, S.C. , WOODMAN, R.F. AND LOWTHER, W.L. ( 1 994).
Role of re-seeding and seedling recruitment for sustainable Lotus corniculatus
based pastures in dry hil l and high country. Proceedings of the New Zealand
Grassland Association 56: 1 39- 142.
GREUB, L.J . AND WEDIN, W.F. ( 1 97 1 ). Leaf area, dry-matter accumulation, and
carbohydrate reserve levels of birdsfoot trefoil as influenced by cutting height.
Crop Science 1 1 : 734-738 .
HARPER, J.L. ( 1 977). Population Biology of Plants. Academic press. New York. p. 84.
HYSLOP, M.G., KEMP, P.D . . AND HODGSON, 1. ( 1 999). Vegetatively reproductive
red clovers (Trifolium pratense L.): An overview. Proceedings of the New
Zealand Grassland Association 61 : 1 2 1 - 1 26.
KEOGHAN, J .M. ( 1 970). The growth of lucerne following defoliation. Thesis of
Doctor of Philosophy. Lincoln College, New Zealand. 3 83 p
LOWTHER, W.L., HOGLUND, J.H. AND MACFARLANE, M.J. ( 1 989). Aspects that
l imit the survival of legume seedlings. In Persistence of Forage Legumes. Edited
by G.C.Marten, A.G. Matches, R.F. Barnes, R.W. Brougham, R.J. Clements and
G.W. Sheath. Proceedings of the Trilateral Workshop, Honolulu, Hawaii. July
1 988. ISBN: 0-89 1 1 8-098-2. pp. 265-275 .
MILLER, J.D., KREITLOW, K.W., DRAKE, c.R. AND HENSON, P.R. ( 1 964). Stand
longevity studies with birdsfoot trefoil. Agronomy Journal 56: 1 37- 1 39.
NELSON, C .J. AND SMITH, D. ( 1 968). Growth of birdsfoot trefoil and alfalfa. Ill .
Changes in carbohydrate reserves and growth analysis under field conditions.
Crop Science 8: 25-28.
PEARS ON, c.J. AND ISON, R.L. ( 1 997). Generation. Chapter 3. Agronomy of
Grasslands Systems. 2nd Edition. Cambridge University Press. pp. 37-59.
PIERRE, J.J. AND JACKOBS, J.A. ( 1 953) . The effect of cutting treatments on
birdsfoot trefoil . Agronomy Journal 45: 463-468.
Chapter Seven 224
REBUFFO, M. and AL TIER, N. ( 1 996). Lotus corniculatus L. LE 65-56 (INIA Draco,
a posteriori). Boletin Interno. (Lotus corniculatus L. LE 65-56. Internal report) . Program a Pasturas. INIA La Estanzuela, Uruguay. 6 pg.
SMITH, D. AND NELSON, CJ. ( 1 967). Growth of Birdsfoot Trefoil and Alfalfa. I.
Responses to height and frequency of cutting. Crop Science 7: 1 30- 1 33 .
TW AMLEY, B .E. ( 1 967). Seed size and seedling vigour i n birdsfoot trefoil. Canadian
Journal of Plant Science 47: 603-609.
TWAMLEY, B .E. ( 1 968). A management study of a birdsfoot trefoil strain trial.
Journal of the British Grassland Society 23: 322-325 .
WHITE, lG.H. AND LUCAS, WJ. ( 1 990). Management of lucerne in the cool season.
Proceedings of the New Zealand Grassland Association 52: 4 1 -43 .
Appendix 225
APPENDIX I
METHOD TO MEASURE TOTAL A V AILABLE CARBOHYDRATES
Extraction with perchloric acid and reaction with anthrone
Clegg ( 1 956) adopted the method of McCready et al. , ( 1 950) including the extraction of
simple sugars with aqueous ethanol and starch with perchloric acid from the residue.
Reagents
Ethanol:
Perchloric acid:
Anthrone reagent:
80% (v/v)
52% (v/v)
1 ml of anthrone (9, 1 0-dihydro-9-oxoanthracene) and 1 0 g of
thiourea was dissolved in one l itre of H2S04 (76%) and stored at
0-4 QC. The colour of reagent increases with the time. Reagent
can be used for two weeks (Southgate, 1 99 1 ).
Glucose standards: A standard glucose solution was diluted to give a serIes of
standards (6.25, 1 2.5 , 25, 50, 100, 200, 400 and 800 J.!g/ml).
Procedures
a. Extraction of sugars (glucose, fructose and sucrose)
0. 1 g of fine dried ground sample was put in a 50 ml centrifuge tube, with two drops of
80% (v/v) ethanol to moisten the sample, then add 2 .5 ml of water and stir thoroughly.
Add 1 2.5 ml of hot 80% (v/v) ethanol, stir for 5 min and centrifuge at 5000 rpm for 1 0
min. Decant the supematant and store, then repeat the extraction adding 1 5 m l of hot
80% (v/v) ethanol. Decant the supematant and combine with the first extraction and
Appendix 226
then remove the ethanol by evaporation at reduced pressure. Filter the mixture and store
in a 1 0 ml volumetric flask adding water to make the volume 1 0 ml. Dilute the solution
to 1 150 dilution, taking 0.2 ml of sample and adding 9.8 of water.
b. Extraction of starch (amylose and amylopectin)
Add 2.5 ml of water to the pellet after the ethanol extraction. Then 3 . 5 ml of perchloric
acid (52%, v/v) and stirring the mixture for 5 min. After that, it was stirred
intermittently for 1 5 min, fol lowing the addition of 1 0 ml of water and centrifuge at
5000 rpm for 1 0 min. Decant the supematant and store in a 50 ml volumetric flask. Re
extract the residue as before and combine the supematant with the first extraction
adding water to make the final volume 50 ml. Filter for anthrone analysis.
c. Analysis of extracts
For blank and each standard solution add 0.25 ml of solution and then 5 ml of anthrone
reagent in tubes with rubber stoppers. For the samples take 0.25 ml of solution, adding
5 ml of anthrone in tubes with rubber stoppers. The content of tubes are mixed and
heated for 1 2 min in a boiling water-bath. Then, tubes are placed in dark for 25 min and
finally the absorbance of the solution is measured at 620 nm.
Appendix 227
APPENDIX II
METHOD TO EVALUATE SOIL SEED RESERVES
The method used is based on direct counting of seeds in soil , fol lowing a technique used
by Prestes ( 1 995).
Procedure
1 . Collection of samples.
Soil cores (22.9 cm2 x 5 cm depth) are taken randomly in the field.
2. Laboratory analysis
The sequence of procedures is represented in Figure 1 .
1 . Samples are hand crumbled
2. Then, samples are sieving in a series of standard sieves from 4.76 to 0.5 mm
aperture
3 . The remaining material i s passed through on air flow to eliminate tiny particles
4. The material is disposed in a becker of 250 ml, adding ethylene cloride (C2CI4). This
is a high density solvent ( 1 .6) separating organic from inorganic material . Finally,
from this supernatant material seeds are hand sorted, separating seeds of Lotus
corniculatus and Trifolium repens and discarding others.
After separation, seeds are counted, weighed and germination tests performed.
Appendix
Hand crumbled
Hand sorting
-
U.S. Standard sieve series
No. 4 - 4,76 mm - Tyler eq. 4 mesh
No. 1 1 - 2 mm - Tyler eq. 10 mesh
No. 35 - 0.5 mm - Tyler eq. 32 mesh
Becker 250 ml, High density solvent Promote separation of organic from
inorganic material
Air Flow
Plate 1 . Soil seed bank analysis
228
